The Eragrostis pectinacea-pilosa
Complex in North and
Central America
(Gramineae: Eragrostoideae)
STEPHEN D. KOCH
URIVERST »
Ay URBANS
4
48
UNIVERSITY OF ILLINOIS PRESS a
ILLINOIS BIOLOGICAL MONOGRAPHS
Volumes 1 through 24 contained four issues each and were available through
subscription. Beginning with number 25 (issued in 1957), each publication is
numbered consecutively. No subscriptions are available, but standing orders
are accepted for forthcoming numbers. Prices of previous issues still in print
are listed below, and these may be purchased from the University of Illinois
Press, Urbana, Illinois 61801. Microfilm and photo-offset copies of out-of-print
titles in the Illinois Biological Monographs are available from University Mi-
crofilms, Inc., 8300 North Zeeb Road, Ann Arbor, Michigan 48106, and the
Johnson Reprint Corporation, 111 Fifth Avenue, New York, New York 10003.
Balduf, W. V. (1959): Obligatory and Facultative Insects in Rose Hips. 12 pls. No. 26. $3.50.
Brandon, Ronald A. (1966): Systematics of the Salamander Genus Gryinophilus. 23 figs.
No. 35. $4.50.
Campbell, John M. (1966): A Revision of the Genus Lobopoda (Coleoptera: Alleculidae) in
North America and the West Indies. 174 figs. No. 37. $5.75.
Jones, George Neville (1968): Taxonomy of American Species of Linden (Tilia). 6 pls. 52
halftones. No. 39. $5.95.
Langebartel, David A. (1968): The Hyoid and Its Associated Muscles in Snakes. 19 figs.
No. 38. $4.95.
Levine, Norman D., and Virginia Ivens (1965): The Coccidian Parasites (Protozoa, Sporozoa)
of Rodents. 2 figs. 48 pls. No. 33. $7.50.
Levine, Norman D., and Virginia Ivens (1970): The Coccidian Parasites (Protozoa, Sporozoa)
of Ruminants. 61 pls. No. 44. $8.50.
List, James Carl (1966): Comparative Osteology of the Snake Families Typhlopidae and
Leptotyphlopidae. 22 pls. No. 36. $13.75.
Morgan, Jeanne (1959): The Morphology and Anatomy of American Species of the Genus
Psaronius. 82 figs. No. 27. $3.00.
Northcutt, R. Glenn (1970): The Telencephalon of the Western Painted Turtle. 19 figs. No.
43. $5.95.
Pinto, John D., and Richard B. Selander (1970): The Bionomics of Blister Beetles of the
Genus Meloe and a Classification of the New World Species. 198 figs. No. 42. $10.00.
Ray, James Davis, Jr. (1956): The Genus Lysimachia in the New World. 20 pls. 11 maps.
Vol. 24, Nos. 3-4. $2.50.
Rominger, James M. (1962): Taxonomy of Setaria (Gramineae) in North America. 6 pls.
15 maps. No. 29. $4.50.
Ross, Herbert H., and William E. Ricker (1971): The Classification, Evolution, and Dispersal
of the Winter Stonefly Genus Allocapnia. 111 figs. No. 45. $8.95.
Sawyer, Roy T. (1972): North American Freshwater Leeches, Exclusive of the Piscicolidae,
with a Key to All Species. 37 figs. No. 46. $8.50.
Selander, Richard B. (1960): Bionomics, Systematics, and Phylogeny of Lytta, a Genus of
Blister Beetles (Coleoptera, Meloidae). 350 figs. No. 28. $4.50.
Selander, Richard B., and Juan M. Mathieu (1969): Ecology, Behavior, and Aduit Anatomy
of the Albida Group of the Genus Epicauta (Coleoptera, Meloidae). 60 figs. No. 41. $5.95.
Stannard, Lewis J., Jr. (1957}: The Phylogeny and Classification of the North American
Genera of the Suborder Tubulifera (Thysanoptera). 14 pls. No. 25. $2.50.
Unzicker, John D. (1968): The Comparative Morphology and Evolution of the Internal Female
Reproductive System of Trichoptera. 14 pls. No. 40. $3.95.
Wrigley, Robert E. (1972): Systematics and Biology of the Woodland Jumping Mouse,
Napaeozapus insignis. 22 figs. No. 47. $7.95.
The person charging this material is re-
sponsible for its return to the library from
which it was withdrawn on or before the
Latest Date stamped below.
Theft, mutilation, and underlining of books
are reasons for disciplinary action and may
result in dismissal from the University.
UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN
L161— O-1096
Digitized by the Internet Archive
In 2011 with funding from
University of Illinois Urbana-Champaign
http://www.archive.org/details/eragrostispectin48koch
The Eragrostis pectinacea-pilosa
Complex in North and
Central America
(Gramineae: Eragrostoideae)
STEPHEN D. KOCH
ILLINOIS BIOLOGICAL MONOGRAPHS 48
UNIVERSITY OF ILLINOIS PRESS URBANA, CHICAGO, AND LONDON
Board of Editors: Donald F. Hoffmeister, Willard W. Payne, Tom L. Phillips, Richard B.
Selander, and Philip W. Smith.
Issued March, 1974.
© 1974 by The Board of Trustees of the University of Illinois. Manufactured in the United
States of America. Library of Congress Catalog Card No. 73-2454.
Library of Congress Cataloging in Publication Data
Koch, Stephen D. 1940-
The Eragrostis pectinacea-pilosa complex in North
and Central America (Gramineae: Eragrostoideae).
(IIinois biological monographs, 48)
Bibliography: p.
1. Love grass. 2. Grasses — North America.
3. Grasses — Central America. |. Title. Il. Series.
QK495.G74K75 584’.93 73-2454
ISBN 0-252-00389-6
CONTENTS
Abstract
Introduction
Materials and Methods
Delimitation of the Eragrostis pecttnacea-pilosa Complex
Taxonomic History of the Compiex
Morphology and Anatomy
Chromosome Number
Ecology
Species Relationships
Description of the Eragrostis pectinacea-ptlosa Complex
Key to Species and Varieties
Eragrostis ptlosa var. pilosa
Eragrostis ptiosa var. perplexa
Eragrostis pectinacea
Eragrostts tephrosanthos
Eragrostis lutescens
Eragrostis franktt
Eragrostis pringlet
Citation of Specimens
Literature Cited
Appendix: List of Collections from Which Chromosome
Counts Were Obtained
Plates
Page
ix
63
67
10.
it.
12.
PLATES
Comparison of the £. pectinacea-ptlosa Complex
and the £. etltanensts Complex
Eragrostis ptlosa var. pilosa
Eragrostts ptlosa var. perplexa
Eragrostis pectinacea
Eragrostis tephrosanthos
Eragrostis lutescens
Eragrostis franktt
Eragrostis pringlet
FIGURES
Section of blade, EF. lutescens (Davis 4131)
Glandular pit, blade of £. lutescens (Davts 4131)
Blade epidermis, F. pectinacea (Barker 1877)
Blade epidermis, £. ctltanensts (Koch 6663)
Section, macrohair and basal pustule; F. pringlet
(Arsene 5441a)
Chromosome number—pollen diameter
Distribution of Z. ptlosa (both varieties)
Distribution of £. pectinacea s.1l.
Uniform culture experiment, "FE. diffusa" extreme
Uniform culture experiment, EF. pectinacea s.s.
extreme
Frequency distribution, summed standardized
values, EF. peettnacea s.l.
Clinal variation in £. pectinacea s.l.
Page
67
68
69
70
71
#2
73
74
36
39
40
Figures (cont'd)
13. Distribution of £. tephrosanthos
14. Distributions of F. frankit, E. pritnglet, and
E. lutescens
MAPS
Distribution of F. ptlosa (both varieties). (Fig. 7)
Distribution of EF. pectinacea s.1. (Fig. 8)
Clinal Variation in EF. pecttnacea s.1. (Fig. 12)
Distribution of £. tephrosanthos. (Fig. 13)
Distributions of EF. frankit, E. pringlet, and E. lutescens.
(Fig. 14)
TABLES
1. Comparison of the £. peettnacea-ptlosa and the
E. ctltanensts Complex
2. Pollen Diameter and Chromosome Number in the
E. peetinacea-ptlosa Complex
3. Important Characters of the Members of the
E. peettnacea-ptlosa Complex
4. Comparison of £. ptlosa s.s., "EF. peregrina," and
"E. perplexa' with Respect to Quantitative
Characters
5. Means, Standard Deviations, and Numbers of
Individuals for Each Square on Fig. 12
6. Source of £. tephrosanthos s.1. Seed Used in the
Uniform Culture Experiment
7. Progeny Test of Plants from Mixed Populations of
E. peettnacea and E. tephrosanthos
Page
46
53
25
33
40
46
53
12
18
27
42
48
50
ABSTRACT
Koch, Stephen D. The Eragrostis pectinacea-ptlosa Complex in North
and Central America (Gramineae: Eragrostoideae). Jllinots Btologt-
cal Monographs, 48. The Eragrostis pecttnacea-ptlosa complex is sep-
arable from the rest of the genus on the basis of six characters: a
non-stoloniferous habit, an annual life cycle, non-disarticulating
rachillas, caryopses which are round or slightly compressed dorsally
in cross section, microhairs in which the basal cell is only 0.8-1.6
times the length of the apical cell, and a mean of 7.0 or fewer sto-
mata per 500-micron-long segment of each abaxial intercostal zone.
As traditionally treated, the complex consists of ten species, #£.
pilosa (L.) Beauv., E. perplexa Harvey, E. peregrina Wieg. (= E£.
multicaulis Steud.), EB. pectinacea (Michx.) Nees, £. diffusa Buckl.,
E. tephrosanthos Schult., HE. arida Hitchc., £. lutescens Scribes 2.
frankit Meyer ex Steud., and #. Pringlei Mattel.
On the basis of examinations of over 2,000 herbarium specimens,
chromosome counts, uniform culture experiments, field observations,
and a statistical analysis of variation, the complex was found to
consist of only seven taxa. These fall into three groups according
to their chromosome numbers and panicle, spikelet, and caryopsis mor-
phology.
A key to the taxa is provided, along with synonymies, descriptions,
distribution maps, discussions, and plates illustrating important fea-
tures for each taxon. No new taxa are recognized, but one new com-
bination is proposed.
ix
= = r F 5 | _ =
' - i ‘ ‘ mu =
= ? o % :
. } s
ay s
'
4
4
\
INTRODUCTION
Eragrostis is a world-wide genus of approximately 300 species.
These are primarily tropical and warm-temperate in distribution,
the genus being especially well developed in tropical and southern
Africa. As treated herein, the Eragrostis pecttnacea-pilosa com-
plex is a group of seven taxa: six species, one of which is com-
posed of two varieties. All but one variety of one species are
restricted geographically to North and Central America, the excep-
tional variety having a pantropical distribution. All are annuals,
three of the taxa are widespread weeds, three are narrow endemics,
and one is infrequent within its range.
For the most part, this paper is adapted from a dissertation ac-
cepted by the University of Michigan in partial fulfillment of the
requirements for the degree of Doctor of Philosophy in 1969. The
original version is available from University Microfilms, Inc., of
Ann Arbor, Michigan. I am especially grateful to Dr. Warren H. Wag-
ner, Jr., chairman of my doctoral committee, for his generous pa-
tience and guidance. I also acknowledge the support of the Herbar-
ium, the Botanical Gardens, and the Department of Botany at the Uni-
versity of Michigan, where most of the research was carried out with
the support of an N.S.F. Graduate Fellowship. In addition, I am in-
debted to the Department of Botany, Duke University, for its support
during the last year of the study.
I wish also to thank Dr. Thomas R. Soderstrom, Curator of Grasses,
for the use of the facilities and extensive collections of the U.S.
National Herbarium and for his advice and encouragement; Dr. Henry
F. Decker, of Ohio Wesleyan University, and Dr. John R. Reeder, of
the University of Wyoming, who allowed the use of some of their un-
published work; Mr. G. William Moore, of North Carolina State Uni-
versity, who assisted in computer programming; Dr. LeRoy H. Harvey,
of the University of Montana, who made valuable comments; and Dr.
xi
ya Eragrostts pecttnacea-ptlosa Complex
John H. Thomas, Curator of the Dudley Herbarium, Stanford Univer-
sity, who sent viable seed of material from California.
For the loan of invaluable materials thanks are also due the
curators of the herbaria listed under Materials and Methods on
page 1 following.
MATERIALS AND METHODS
Approximately 2,000 herbarium specimens were examined, in addition
to my own collections from Arkansas, California, Florida, Georgia,
Illinois, Kansas, Louisiana, Massachusetts, Michigan, Missouri,
North Carolina, Ohio, Oklahoma, Tennessee, and Texas in the United
States, and Jamaica and Costa Rica. The majority of the herbarium
specimens were lent by the U.S. National Herbarium, but collections
from other herbaria were examined also. Material was borrowed from
the following institutions; their textual abbreviations according
to Lanjouw and Stafleu (1964) appear in parentheses:
Botanische Staatssammlung (M)
British Museum (Natural History) (BM)
Canadian Department of Agriculture, Phanerogamic Herbarium
(DAO)
Duke University (DUKE)
Laboratoire de Phanérogamie, Muséum National d'Histoire
Naturelle (P)
Philadelphia Academy of Natural Sciences (PH)
Rancho Santa Ana Botanic Garden (RSA)
Texas Technical College (TTC)
Tracy Herbarium (TAES)
United States National Herbarium (US)
University of Kansas (KANU)
University of Michigan (MICH)
Uniform culture experiments were performed by planting, at the
same time in a greenhouse, caryopses of the plants to be compared
and allowing them to grow to maturity on the same bench. Plants
for progeny tests were grown in an experimental garden. In all
cases, examples of mature plants were pressed. These voucher spec-
imens are deposited in either the University of Michigan Herbarium
or the Herbarium at North Carolina State University. No attempt
2 Eragrostts peettnacea-ptlosa Complex
was made to control environmental conditions precisely, so compari-
sons are valid only among plants included in a particular test.
Observations of leaf blade epidermal anatomy were made on all
members of the complex and on some species not included in the
group. Leaf blade cross sections were examined only to determine
the structure of glandular pits and cushioned macrohairs. Epider-
mises were prepared according to the method outlined by Metcalfe
(1960, p. lx-lxi). Cross sections were made by hand. In all cases
material was stained in Delafield's haematoxylin and safranin.
Pollen was stained with cotton blue in lactophenol for measure-
ment and determination of viability. The diameters of a random
sample of 15 grains per collection were determined. Whenever pos-
Sible, at least 15 collections from each taxon were sampied; when
fewer than 15 collections were available, all were used. The mean
diameter was determined for each pollen collection, and from these
a "mean of means'' and standard deviation of means were calculated
for each taxon.
Chromosome counts were made from pollen mother cells by the stan-
dard acetocarmine squash method. An effort was made to determine
the chromosome numbers of a relatively large sample from each taxon,
and to select the collections from as much of the range of the tax-
on as possible. In taxa in which living material was limited, the
chromosome number of every collection was determined. All of the
collections from which counts were obtained are listed in the Ap-
pendix (p- 63).
DELIMITATION OF THE ERAGROSTIS PECTINACEA-PILOSA COMPLEX
The Eragrostis pectinacea-pilosa complex can be separated from the
rest of the genus by six characters: (1) an annual life cycle, (2)
a non-stoloniferous, upright habit, (3) non-disarticulating rachil-
las, (4) caryopses which are rounded in cross section, being neither
longitudinally grooved nor strongly flattened, (5) bicellular micro-
hairs in which the length of the basal cell is 0.8-1.6 times the
length of the apical cell, and (6) a mean of 7.0 or fewer stomata
on a 500u-long segment of abaxial intercostal zone.
Delimitation of Complex 3
So defined, the F. pectinacea-pilosa complex in North and Central
America includes the following ten commonly recognized species:
E. artda Hitche. E. pectinacea (Michx.) Nees
E. dtffusa Buckl. E. perplexa Harvey
E. frankti Meyer ex Steud. E. ptlosa (L.) Beauv.
E. lutescens Scribn. E. prtnglet Mattei
E. multicaults Steud. E. tephrosanthos Schult.
In circumscribing the £. pectinacea-pilosa complex, the most dif-
ficult problem faced was that of the disposition of F. barreltert Da-
veau, FE. poaeoides Beauv. ex Roem. and Schult., and £. berterontana
(Schult.) Steud. They differ from the £. pecttnacea-pilosa complex
only in the ratio of the basal to apical cells of the microhairs and
the mean number of stomata per 500u of intercostal zone on the abaxial
leaf surface. These characters, however, correlate well with differ-
ences in the general appearance of these two groups of species, es-
pecially that of the panicle and spikelets. Also, they are apparently
part of a closely related group of species that includes F. etlianensts
(All.) Lutati, a species with disarticulating rachillas. For conveni-
ence, this group is called the £. ciltanensis complex. This assem-
blage seems not to be especially closely related to FE. pectinacea, E.
pilosa, and their allies. The difficulty in establishing this appar-
ent lack of relationship lies in the fact that most of the characters
separating the two groups involve differences in degree rather than in
kind, and, as such, do not lend themselves to concise statements. The
two complexes are compared in more detail in Table 1 and Plate I.
Based on the differences in spikelet and panicle morphology listed
in Table 1, both Bentham and Hooker (1883) and Pilger (1956) placed
the species belonging to these two groups in different subsections of
section Eragrostis (= section Pteroessa Doell). The £. peettnacea-
pilosa complex was placed in subsection Leptostachyae Nees and the £.
ciltanensts complex in subsection Megastachya Benth.
Harvey's treatment (1948) of the species in these two groups dif-
fers from both of these and from mine. He placed both of them in his
section Eueragrostis (sensu Boissier, 1884), which is divided into ten
subsections. His subsection Cilianenses included the four species in
4 Eragrostts pecttnacea-pilosa Complex
TABLE 1
COMPARISON OF THE F. PECTINACEA-PILOSA
AND THE £. CILIANENSIS COMPLEX
Character
Panicle (Plate I, a & b)
Appearance
Branches
Spikelet (Plate I, c & d)
Shape in cross section
Lemma apex
Lemma lateral nerves
Lemma opacity
Glands (Plate I, e & f)
Frequency of
characteristically
glandular species
Distribution on culm
and panicle axis
Type
Epidermal anatomy*
(Figs. 3 & 4)
Microhair basal cell
length/apical cell
length ratio
Stomatal density~
a
Data calculated from
abaxial epidermis of one
Peaced on a sample of
“Measured by counting
of each intercostal zone
E. pecttnacea-pilosa
Complex
Relatively dense
Relatively long, capil-
lary to filiform; sev-
eral secondary and
sometimes tertiary
branches
Laterally compressed;
dorsal sides V-shaped
Acute to acuminate
Moderately conspicuous
to inconspicuous
Translucent
2 of 7 species
Scattered; never
coalesced
Glandular pits with
flush margins
0.8-1.7
~ I
Il
RET 7 «i
*
Il
E. ctltanensts
Complex
Relatively open
Relatively short,
stout; few sec-
ondary branches,
tertiary bran-
ches rare
Weakly laterally
compressed; dor-
sal sides rounded
Obtuse to acute
Conspicuously
raised
Opaque
3 of 4 species
Restricted to defin
ite rings beneath
nodes; usually
coalesced
Glandular pits with
flush margins, an
crateriform gland
with raised margi
tO Sate
10.4-11.1
*
Wl
random samples of the characters on the
leaf typical of each taxon.
ten hairs per leaf.
the number of stomata in a segment 500U long
across a leaf.
Taxonomic History of Complex 5
the Z. etlianensts complex, in addition to £. lutescens, which, on the
basis of the characters just discussed, is clearly related to F. pec-
tinacea and FE. ptlosa. His subsection Pectinaceae included four of the
remaining six taxa of the #. pectinacea-pilosa complex, along with sev-
eral annual species which have caryopses with a deep groove in their
dorsal surfaces, such as £. mexicana (Hornem.) Link and £. oreutttana
Vasey. Caryopsis shape has been largely ignored as a character in £ra-
grostis, but it is effective in separating a great many species, both
annual and perennial, and correlates well with other, less easily de-
scribed features. It holds promise in helping to define natural sub-
generic groups. Because of the apparent importance of this character,
those species with grooved caryopses are excluded from the F. peetitnacea-
ptlosa complex.
Harvey's subsection Capillares, also of section Fueragrostis, included
E. franktt and E. pringlet, along with EF. glomerata (Walt.) L. H. Dewey
and F. capillaris (L.) Nees. Subsection Capillares is apparently an un-
natural assemblage, as it is merely a group of annual species having long,
usually spreading pedicels, and small spikelets. F. captllaris has
grooved caryopses and is virtually identical to one- or two-year-old spe-
cimens of £. intermedia Hitchc., a perennial which also has grooved car-
yopses. It appears that F. capillaris has its affinities with £. inter-
media and its grooved-seeded, perennial allies. The disarticulating
rachillas and spikelet morphology of EF. glomerata iead me to believe
that this species is related to those in section Cataclastos Doell—for
example, Z. ciliaris (L.) R. Br. and £. tenella (L.) Beauv. ex Roem. and
Schult. This opinion is supported by Pilger (1956, p. 11), who placed
E. glomerata in section Cataclastos. The remaining two species, FE. fran-
kit and E. pringlet are members of the FE. pecttnacea-ptlosa complex.
TAXONOMIC HISTORY OF THE COMPLEX
Knowledge of the Eragrostis pectinacea-ptlosa complex grew slowly until
1900. Linnaeus had named but one species, #. ptlosa (as Poa pilosa),
and it was not until 1803 that Michaux named the second. During the
nineteenth century, numerous authors treated the complex, but always as
part of a treatment of the whole genus. Noteworthy among these are
Muhlenberg (1817), Nuttall (1818), Schultes (1824), Gray (1848, 1856),
6 Eragrostis pecttnacea-ptlosa Complex
Steudel (1854), and Beal (1896). The work of these authors has been
reviewed by A. S. Hitchcock in his various floras (1909, 1913, 1930,
1935, 1936, 1951) and in his published studies of some of the earlier
herbaria (1905, 1932).
By 1900, six of the ten traditionally recognized species of the F.
pectinacea-pilosa complex had been named, but considerable confusion
prevailed about the proper application of some names, and others were
incorrect according to current nomenclatural concepts. Through his study
of the types of American grass names, A. S. Hitchcock determined the
correct names for these taxa. In addition, his flora brought together
all of the new species that had been described since Beal's Grasses of
North America (1896). The first edition of Hitchcock's Manual (1935)
recognized nine of the species in the complex. Harvey (1948) added
the tenth, #. perplexa. All subsequent authors, including Harvey, -
have followed Hitchcock's treatment.
MORPHOLOGY AND ANATOMY
The taxonomically important variation in the morphology of the Fra-
grostis pectinacea-pilosa complex is summarized in Table 3 (p. 18).
Among grasses, one of the unusual features of Fragrostis is that
many of its members bear glands. Plants of two of the seven taxa of
the £. pectitnacea-ptlosa complex are characteristically strongly glan-
dular, and at least a few specimens referable to each of the other
taxa bear glands. In the taxa characterized by the glandular condi-
tion, #. lutescens and FE. pilosa var. perplexa, the glands are scat-
tered over the whole plant, being most dense on the culms, lower pan-
icle axis, sheaths, and lower portion of the abaxial surfaces of the
blades. Among the taxa characterized by the lack of glands, excep-
tional glandular individuals are most often encountered in EF. fran-
kit and its close relative £. pringlet. Gland distribution in £.
frankii varies widely; the plant may be entirely eglandular, or may
have glands in all parts. In £&. pringlet the plants are eglandular
except for the pedicels, which always bear a glandular ring at or
near their midpoints. Glands are relatively rare in the rest of the
taxa. When they do occur, they are usually few in number, being
Morphology and Anatomy 7
found beneath several of the culm nodes and around one of the lowest
two panicle nodes, or only around one of the lowest two panicle nodes.
These glands have been called glandular pits (Harvey, 1948). Under
magnifications of 10x to 20x, they appear to be shallow depressions of
varying size, with diameters in the order of 0.05-0.10 mm, round or el-
liptic in outline, and with a flush to slightly raised margin (Plates
IIlId and VId). They always occur over veins. Anatomically, these glan-
dular pits closely resemble the epidermal glands found in other species
of Eragrostis (Nicora, 1941). They are made up of thin-walled, infla-
ted epidermal cells which stain purple with Delafield's haematoxylin
and safranin, whereas the surrounding epidermal cells stain red. The
glandular layer is only one cell thick; it does not interrupt the fi-
bers of the bundle cap just below the epidermis (Figs. 1 and 2).
Nicora (1941) examined the exudate of the glands of several spe-
cies of Eragrostis. She found that it consists of a white, crystal-
line substance and a viscous, yellow fluid. The latter was found to
contain sugars and oils. In £. ctlianensis, a common and highly glan-
dular species, this secretion gives the plant a strong phenolic odor,
somewhat reminiscent of crushed cockroaches. The function of this
scent is unknown, although it is tempting to believe that it acts as
an animal repellent. This odor has not been detected in members of
the £. peetitnacea-pilosa complex, but the only two taxa that have
abundant glands are rare and were not seen in the field.
Cushioned macrohairs are found in the closely related EF. frankit
and E. pringlet. These hairs are visible to the naked eye, unicel-
lular, relatively stiff, thick-walled, and have a bulbous base. More-
over, the bases are sunken into pustules which are made up of special-
ized epidermal cells (Fig. 5). According to Metcalfe (1960, p. xliii),
macrohairs which are sunken into the epidermis are characteristic of
species native to warm climates, while superficial macrohairs are
most often found in species from temperate climates. This suggests
that £. frankii, the only temperate member of the complex which has
these cushioned macrohairs, may have its evolutionary affinities in a
more tropical region than it now occupies.
8 Eragrostis pecttnacea-ptlosa Complex
Fig. 1. Cross section of blade of F. lutescens (Davis 4131) through
a vascular bundle and glandular pit (stippled). Mesophyll not shown.
AB—abaxial epidermis; AD—adaxial epidermis; BC—bundle cap; VB—
vascular bundle.
| 25p ;
Fig. 2. Surface view of a glandular pit (stippled) on abaxial sur-
face of midrib of a blade of EF. lutescens (Davis 4131).
Morphology and Anatomy
—- <—_ S< —_ 2
SK M Cee.
SS ee | :
| ,
ee
8 5 Wasa FS ea. Us eee 1
Ly ae om te eee
see eaten fae mere Pe I Tease Up eS
Sa ee
(Toe TT EO EEF
Geter PEER
Fig. 3. Abaxial blade epidermis of £, peettnacea (Barker 1877).
C—costal region; I—intercostal region; M—microhair; S—stomatal
apparatus.
en aetaeecres
Mo er —— os - Mt)
ANP TT ] 50
a TD OPTS ITT
Ca rh
5 Op
Fig. 4. Abaxial blade epidermis of F. ciltanensts (Koch 6663)
10 Eragrostis pectinacea-pilosa Complex
25p
Fig. 5. Section through a macrohair and its basal pustule; JL.
pringlet (Arsene 544la).
Chromosome Number 11
The leaf blade epidermises in the members of the £. pecttnacea-
pilosa complex were found to be virtually identical except that some
of them have glands or cushioned macrohairs. Fig. 3, a drawing of ‘
the abaxial epidermis of £. pectinacea, is typical of members of
this group. Cross sections of leaf blades were not examined except
to determine the structure of the glandular pits and cushioned mac-
rohairs.
Reeder (personal communication) has examined the anatomy of the
embryos of thirty-one species of Eragrostis, including three which
are members of the £. pectinacea-pilosa complex. He found them to
be identical to that of F. pecttnacea, which is illustrated in his
paper (Reeder, 1957).
CHROMOSOME NUMBER
Chromosome number was found to be extremely useful in treating the
Eragrostis pectinacea-pilosa complex. In addition, it is believed
that it will be very helpful in dealing with Fragrostis as a whole.
Table 2 lists the chromosome numbers and the number of collec-
tions on which these are based for each taxon. Documentation for
these numbers is in the Appendix. Plates IId, IVd, Vd, and VIId
are photographs of meiotic chromosomes of members of the complex.
In general, the numbers listed in Table 2 agree with literature re-
ports (Tateoka, 1965; Ono and Tateoka, 1953; Reeder, 1971; Gould,
1958 in part, 1965). Two reported numbers, however, are at vari-
ance with mine. Gould (1958) obtained a gametic number of 20 in
one collection of the normally hexaploid £. pectinacea. I have ex-
amined the voucher specimen (Gould 7623, TAES), and it appears to
be identified correctly. Its mean pollen diameter is 27.4u, between
the tetraploid and hexaploid groups (see below). Bowden and Senn
(1962) reported 2n=60 in a collection of the usually tetraploid F.
pilosa var. pilosa (listed as "FZ. pilosa'') from Argentina. Plants
grown from seed from the original collection (Senn 4129, DAO) were
revealed to be specimens of £. virescens Presl.
Data in Table 2 show that there are three ploidal levels in the
complex. Two of the taxa are hexaploids (2n=60), one is tetraploid
12 Eragrostis pectinacea-ptlosa Complex
TABLE 2
POLLEN DIAMETER AND CHROMOSOME NUMBER
IN THE &. PECTINACEA-PILOSA COMPLEX
Chromosome number Pollen diameter (ne
Taxon yw? 2n° nP x= s— range —
= x _—s
E. pectinacea 27 60 30: 31.9. 3.0 -2659=35.6
E. tephrosanthos 6 60 15 -30e1. “225 2Z7el=33s8
E. lutescens A 33.0 2.8 31.0=3528
E. ptlosa var. ptlosa ti 40 30. 26.6. 2.3: 22.8=31.6
E. ptlosa var. perplexa 5: 2762 266° 2505-2942:
E. franktit 6 40 15 24.9 2.2 22.6-28.1
E. pringlet & 23.1. 2.0. 20.6-26.1
E. frankit 1 80 1 29.5
*all pollen parameters derived from means of 15 grains per plant.
Piambee of collections sampled for each taxon.
c .
Somatic chromosome number.
(2n=40), and one has both tetraploid and octoploid (2n=80) members.
There are no diploids (2n=20); in fact, diploids seem to be rela-
tively infrequent in the genus as a whole, to judge from published
reports. Within each taxon, the chromosome number was found to be
uniform, despite the rather wide distribution of some of the taxa.
The exception to this is the tetraploid £. frankii, in which one
octoploid population was found.
No abnormal meioses were encountered in any of the taxa examined,
and pollen viability, as judged by morphology and stainability with
cotton blue, was approximately 90%.
5
Chromosome Number
"(@T *d) 7 ATqeL worzz Bled
*Iaqunu OTJeWOS 9Yy} ST Joqunu swosowoIy)
xX
*3u0T (=sz) SUOTJETAOP pAePpueIS OM} ST BUTT TedTIAeA £(-X) UOXe] |9Y. AOJ JaezeWeTp uaT{Tod uesew
aud ST oUuTT Te.UOZTAON
wl
soy.ups
-oayda. *i_
09
*‘Joqowetp uetTtTod pue Aequnu suosowo1zyo usemjzeq dtysuotjepTey
(ptotdeizja3) vxe7zdiead *1ea pso7id *1eA
peavuiz0ed *qy 2re7bu1ad *q = 12yUDIf “7 psojzid *q psojid °*4
09 74 Ov _ Ov
(protdoqs0)
11yUundf *“q@ suaosazn7
08 =
a se
oe
Le
je
j=
oO
°9g *3TY
uoxe]
aequnu swosowoi1yuy)
“GE
“Of
(1)
Jazewerp
ueaT [Od
14 Eragrostts pectinacea-ptlosa Complex
Table 2 also lists the mean pollen diameter for each taxon (xX),
its standard deviation (Sz); and the range of means for each. Fig.
6 was prepared from Table 2 to show more clearly the relationship
between pollen diameter and chromosome number. It can be seen that
the members of the complex fall into two groups, those with a chro-
mosome number of 2n=40 and smaller pollen, and those with 2n=60 and
larger pollen.
In Fig. 6 are plotted the pollen diameters of the three taxa for
which no cytological material was available. On the basis of pollen
diameter, FE. pringlei and F. pilosa var. perplexa are expected to
have a chromosome number of 2n=40, while £. lutescens should have
2n=60. The use of pollen diameter to infer chromosome number is
limited, however, by the fact that the pollen diameter of the octo-
ploid population of £&. frankii falls within the range of those of
the hexaploids.
ECOLOGY
The members of the complex are annuals which grow best on highly
disturbed, damp to wet, sandy soils. With the exception of £. pt-
losa var. pilosa, which is pantropical, all are restricted to North
and Central America. &. pectinacea, E. tephrosanthos, and E£. ptlosa
var. pilosa are weeds; £. pringlei, E. lutescens, and FE. pilosa var.
perplexa are rare plants; and £. frankii, although moderately com-
mon, is not weedy.
Of the seven taxa in the £. pectinacea-pilosa complex, living
material was available for only four: #£. pilosa var. pilosa, E.
pectinacea, E. tephrosanthos, and £. frankii. Consequently, the
following statements are based on observation of these four, al-
though no evidence was encountered which would suggest that extrap-
olation to the other three is invalid.
The members of the complex have a life cycle, from seed to seed,
of about 60 days in the greenhouse. Perennation under cultivation
is rare; usually the plants die after most of the panicles are ma-
ture, regardless of the temperature and water supply. Whether per-
ennation occurs in nature is unknown. Water deprivation can speed
Ecology 15
up the life cycle, but the plants are severely stunted, and produc-
tion of viable caryopses is then very poor.
The caryopses have a dormancy requirement, as is commonly the
case with grasses. No germination will occur immediately after they
are released from the plant, but after six to eight months, close to
100% germination within about a week is obtained. Scarification
(most easily accomplished by soaking the caryopses and removing the
apical end with a razor blade) can increase the amount of germina-
tion at any time prior to six months, although the germination thus
achieved is always considerably less than 100%. Good germination is
maintained for five to seven years, as caryopses from herbarium spec-
imens this old usually germinate readily. There is no noticeable
difference with respect to these factors among species or among col-
lections, although caryopses from both tropical and temperate collec-
tions were grown.
The seedlings, which appear in late spring, are intermediate be-
tween the panicoid and festucoid types. The panicoid type has a
first leaf which is horizontal, short and broad, whereas the festu-
coid type has a first leaf which is erect, long, and narrow (Prat,
1936). In the Fragrostts pectinacea-ptlosa complex, the first leaves
vary from horizontal to ascending, and are usually narrowly elliptic
in outline.
Panicles begin to appear about a month after germination and are
produced until the plant dies. It is curious that, even in the
greenhouse, the plants die with half-emerged panicles on them.
In Eragrostts pectinacea, the only species observed closely,
flowering is a brief process, lasting about 90 minutes from the
first noticeable separation of lemma and palea to floret closing.
During this time, the styles and filaments elongate rapidly, finally
becoming exserted. The anthers at the time of dehiscence are so
close to the plumose stigmas that self-pollination is inevitable.
Soon after anther dehiscence, the lemma and palea begin to close,
often leaving one or more anthers or stigmas, or both anthers and
Stigmas, partially or wholly exserted from the floret.
The three most widespread taxa, F. pectinacea, E. ptlosa var.
16 Eragrostis pecttnacea-pilosa Complex
pilosa, and FE. tephrosanthos are either apomictic or self-fertile.
This was demonstrated by enclosing immature panicles in glassine
bags just as they were emerging from the sheath. The bags were left
in place until the panicles matured, producing, in every case, abun-
dant, well-formed caryopses. The possibility of outcrossing under
natural conditions is not precluded by this experiment, but, if it
does occur, its frequency must be rather low (barring the presence
of an internal mechanism favoring the development of alien pollen)
because self-pollination is made a virtual certainty by the manner
in which the anthers dehisce.
Seed production is copious. The caryopses of all species are
about 0.5 mm long and have no special adaptations for dispersal.
Transport with the upper layer of soil is probably the most impor-
tant natural method of distribution. For the weedy taxa, dissemin- |
ation by commercial traffic is especially effective. It is common
to find a taxon growing outside of its normal range, but restricted
to railroad or truck yards, where the plants apparently are unable
to become established beyond the immediate site of introduction.
SPECIES RELATIONSHIPS
On the basis of their geography, chromosome number, and morphology,
the species of the Eragrostis pectinacea-pilosa complex can be di-
vided into three groups. Eight characters are particularly useful
for this differentiation: (1) length of the sheath with respect to
its internode, (2) panicle morphology, (3) arrangement of the pri-
mary branches at the two lowest panicle nodes, (4) spikelet length,
(5) palea persistence, (6) glume length, (7) caryopsis shape, and
(8) chromosome number. On the basis of these and certain other
characters, the members of the group are compared in Table 3 (p. 18).
The two varieties of £. pilosa make up the first group. Within
the complex, they are the only taxa that have short glumes, the
first glume being less than half the length of the lowest lemma; de-
ciduous paleas; and primary panicle branches usually whorled at one
of the two lowest nodes.
There is little question that the two are closely related and
Species Relationships 17
that #. pilosa var. perplexa is only a minor evolutionary offshoot
of #. ptlosa var. ptlosa. The only morphological difference between
the two is in the distribution of glands, £. pilosa var. perplexa
always having glandular pits scattered over the whole plant, and £.
ptlosa var. ptlosa being eglandular, or, at most, having a few glands
beneath one of the two lowest panicle nodes. Geographically, the
distributions are nearly contiguous, £. ptlosa var. pilosa extending
into eastern Texas, #. ptlosa var. perplexa occupying the Plains to
the north and west of this area. Indeed, they may actually be con-
tiguous, the rarity of F. ptlosa var. perplexa or the lack of col-
lecting—or perhaps both factors—causing the apparent disjunction.
The second group consists of three closely related species, £.
pecttnacea, E. tephrosanthos, and E. lutescens. They all have a
chromosome number of 2n=60, a unique character in the complex. The
relationship between £. pectinacea and FE. tephrosanthos is very
close, the only morphological difference between them being whether
the pedicels borne laterally on the panicle branches are appressed
to ascending (£. pectinacea) or spreading to divaricate (£. tephro-
santhos). Geographically their ranges overlap broadly in the south-
western United States and Mexico, but are separate in the eastern
United States and the West Indies. Despite their sympatry and mor-
phological similarity, no evidence of hybridization has been found
(see discussion of #£. tephrosanthos, p. 44).
E. lutescens, the third member of this group, differs from the
others in having glands scattered over the whole plant and its pan-
icle branches closely ascending or appressed to the panicle axis.
Geographically, it is completely isolated from F&. pectinacea and E£.
tephrosanthos. The evolutionary affinity of #. Zutescens 1s probab-
ly with #. pecttnacea because it shares the characteristic of ap-
pressed pedicels with this species.
The third group includes Eragrostis frankii and EF. prtnglet.
These two species are unique in that they have short spikelets; el-
lipsoid, compactly branched panicles; caryopses which are broadly
ellipsoid to globose and smaller than those of the other taxa; and
sheaths which exceed the length of their internodes. The two spe-
18
IMPORTANT CHARACTERS OF THE MEMBERS OF THE
Character
Sheath length:
internode
length
Macrohair
distribution
Glands
Panicle
morphology
Primary branch
arrangement at
lowest two
panicle nodes
Pedicel angle
Florets/
spikelet
Palea
persistence
First glume
length: lowest
lemma length
Caryopsis
shape
Caryopsis
length (mm)
Caryopsis
color
Somatic
chromosome
number
Eragrostis pectinacea-ptlosa Complex
TABLE 3
E. PECTINACEA-PILOSA COMPLEX
E. ptlosa var.
ptlosa
Plate II
<1
Sheath margin
apex only
Rare
Ellipsoid to
ovoid, open
Whorled or
opposite
Appressed to
spreading
(6-) 8-10
Deciduous
Lamp st-ly2)
Pyriform to
ovoid
0.5-0.9
Brown
2n=40
E. ptlosa var.
perplexa
Plate III
<l
Sheath margin
apex only
Abundant
Ellipsoid to
ovoid, open
Whorled or
opposite
Appressed to
spreading
(6-) 8-10
Deciduous
Lf4-173(-1/2)
Pyriform to
ovoid
0.8-1.0
Brown
2n=40"
E. pecettnacea
Plate IV
<1
Sheath margin
apex only
Rare
Ovoid to pyra-
midal, open
Alternate or
opposite
Appressed
(6-) 8-10
Persistent
1/2-3/4
Pyriform
0.5-1.1
Brown
2n=60
“Inferred from pollen diameter.
E. tephrosanthos
Plate V
<1
Sheath margin
apex only
Rare
Ovoid to
pyramidal, open
Alternate or
opposite
Spreading
(6—). 8-10
Persistent
1/2-3/4
Pyriform
0.6-1.1
Brown
2n=60
Species Relationships
Table 3 (cont'd)
E. lutescens
Plate VI
si
Sheath margin
apex only
Abundant
Branches ap-
pressed to
axis
Alternate
Appressed
(6=) °S=10
Persistent
1/2-3/4
Pyriform
ca. 0.8
Brown
2n=607
E. frankit
Plate VII
>1
Sparse on
sheaths
Infrequent
Ellipsoid,
compactly
branched
Alternate
Spreading
3-4 (-6)
Persistent
3/4-1 1/4
Broadly ellip-
soid to globose
0.4-0.6
Brown
2n=40
BS |
E. pringlet
Plate VIII
>1
Abundant on
sheaths & blades
Pedicels only
Ellipsoid,
compactly
branched
Alternate
Spreading
3-4 (-6)
Persistent
1/2-3/4
Globose
Whitened
2n=40"
20 Eragrostts pectinacea-ptlosa Complex
cies differ only in the minor characters of the whitening of the
caryopsis and hairiness of the sheaths and blades. They are clear-
ly closely related, but geographically they are widely separated.
E. frankit occurs in the eastern deciduous forest of the United
States; E. pringlezt is known from only two localities in central
Mexico.
At present, the evolutionary relationships among these three
groups are not clear. This is because, in order to speculate about
evolutionary pathways, it is necessary to decide which character
states are primitive within the whole genus, and to investigate the
genera which are closely related to Eragrostis. Such studies are
beyond the scope of this endeavor.
DESCRIPTION OF THE ERAGROSTIS PECTINACEA-PILOSA COMPLEX
Annuals. Culms one to several, erect or ascending, sometimes branch-
ing at the lower nodes, at most the lowest 2 or 3 nodes decumbent
and rooting. JLtgule a dense row of short, white hairs. Blades lin-
ear to tapering from a broad base. Panicles narrowly ellipsoid to
pyramidal; primary branches spreading or appressed to the axis, cap-
illary or filiform, often flexuous, the longer usually with secondary
branches, less frequently with tertiary branches. Pedicels capillary,
often flexuous, those arising directly from the panicle axis nearly
equaling or exceeding the length of their spikelets. Sptkelets nar-
rowly ovate to ovate, the base obtuse, moderately compressed, rachil-
la not disarticulating at maturity. Gluwmes ovate to lanceolate,
acute to acuminate, membranaceous, l-nerved, deciduous. Lemmas glab-
rous, ovate, acute to accuminate, membranaceous to chartaceous, mod-
erately compressed, usually weakly keeled, lateral nerves inconspic-
uous to moderately conspicuous, pellucid to opaque, imbricate, decid-
uous, diverging from the rachilla at an angle of 60° or less. Paleas
membranaceous, arched, 2/3 the length of, to nearly equaling, the
lemmas. Caryopses in cross section round, elliptic, convex-triangu-
lar, or slightly flattened dorsally, never with a deep longitudinal
groove in the dorsal side and not strongly flattened laterally; em-
bryo 0.3-0.8 times the length of the caryopsis, its long axis parallel
Key to Species and Varieties 21
to that of the caryopsis. Microhairs clavate, apical cell cylindri-
cal, rounded, basal cell 0.8-1.6 times the length of the apical cell.
Stomata of abaxial blade surface mostly in one, less often two, rows
in each intercostal zone, mean number of stomata per 500u-long sec-
tion of each abaxial intercostal zone 4.7-7.0.
KEY TO SPECIES AND VARIETIES
1. Panicle branches closely ascending or appressed to the axis at
maturity; sheaths and abaxial surfaces of blades with glandular
Dit ovdsyaeersexeeGns Mee weees nes Seas eawer een 5. E. lutescens
1. Panicle branches spreading at maturity; sheaths and abaxial
blade surfaces with or without glandular pits.
2. Larger spikelets with fewer than 6 florets; panicles com-
pactly branched, ellipsoid, the middle branches longest;
lateral pedicels on panicle branches spreading.
3. Caryopses chestnut-brown, not whitened; blades and
sheaths glabrous to sparsely pilose... 6. E. frankit
3. Caryopses strongly whitened; blades and sheaths densely
HEPSUCN Pdd aa Soe ta athens cond Baas Js Be prengtet
2. Larger spikelets with more than 5 florets; panicles open,
ovoid or pyramidal, the lower branches longest; lateral
pedicels on panicle branches appressed to spreading.
4. Sheaths and blades with few to many scattered glandular
DLUSsacseueuSiieedasns oes 2. E. ptlosa var. perplexa
4. Sheaths and blades eglandular.
5. Paleas deciduous, at least on some spikelets;
first glumes less than half as long as the lowest
lemmas; panicle branches at one of lowest two
nodes whorled or, infrequently, fascicled or
PAIVEds iwwaveavses eae se 1. E. ptlosa var. ptlosa
5. Paleas persistent; first glumes more than half as
long as the lowest lemmas; panicle branches at low-
est two nodes solitary or paired.
6. Lateral pedicels on panicle branches appressed,
rarely diverging as much aS 20°....sesseeeeeee
22 Eragrostis pectinacea-pilosa Complex
6. Lateral pedicels on panicle branches, at
least in part, spreading to divaricate at
maturity.
7. Middle sheaths mostly shorter than their
internodes; Caryopses PYTiformss<s ss eee'es
GDA Out hae pea wwe ed 4. E. tephrosanthos
7. Middle sheaths mostly longer than their
internodes; caryopses globose to broadly
CILIPSOLC 4 6 seas aaa etek ere 6. &. frankit
]. ERAGROSTIS PILOSA (Linnaeus) Beauvois var. PILOSA
Poa pilosa Linnaeus, 1753, p. 68.
Eragrostis pilosa (Linnaeus) Beauvois, 1812, p. 71.
Poa tenella sensu Pursh, 1814, p. 80. [Non Linnaeus (1753).]
Eragrostts filtformis Link, 1827, p. 191. [Interpreted from de-
scription and reduction of Poa tenella sensu Pursh to synonymy. ]
Poa linkit Kunth, 1829, p. 113. [Based on Eragrostis filtformis
Link: |
Eragrostis linkit (Kunth) Steudel, 1854, p. 273.
Eragrostts multtcaulis Steudel, 1854, p. 426. [Interpreted from
description. ]
Eragrostis pilosa var. damtenstana Bonnet, 1881, p. 412. [Lecto-
type: Damtens (Exsicc. Société Dauphinoise 3100). (P). Selec-
ted from two syntypes by Stephen D. Koch. ]
Eragrostis tnsconsptcua [sic] Hortus ex Bonnet, 1881, p. 412. [Pro
synonymo £. pilosa var. damtensiana Bonnet. ]
Eragrostis ptlosa var. condensata Hackel, 1902, p. 13. [Holotype:
A. Kneucker Gram. Exsicc. 115. (W). Paratype and fragment of
holotype seen at US. ]
Eragrostis pilosa ssp. damtenstana (Bonnet) Thellung, 1907, p. 438.
Eragrostis ptlosa ssp. danienstana var. condensata (Hackel) Thel-
lung, 1907, p. 439.
Eragrostis peregrina Wiegand, 1917, p. 95. [Based on £. ptlosa var.
condensata Hackel. ]
Eragrostis pilosa var. pilosa 23
Eragrostis danienstana Bonnet ex Thellung, 1928, p. 323.—Bonnet,
1881, p. 412. [Pro synonymo £. pilosa var. damtenstiana
Bonnet. |
Eragrostis damiensiana var. laxtor Thellung, 1928, p. 327. [Based
on £. pilosa ssp. danienstana (Bonnet) Thellung, excluding var.
condensata (Hackel) Thellung. |
Eragrostis damtenstana var. condensata (Hackel) Thellung, 1928, p.
528.
Culms 8-66 cm tall, very rarely with a few glandular pits be-
low one or two of the nodes. Sheaths eglandular; from shorter than,
to equaling, their internodes; glabrous except for a sparse tuft of
white hairs at the apex of the margins, or entirely glabrous. Blades
glabrous, eglandular, the uppermost 2-20 cm long, 1.0-3.5 mm wide at
the widest point. Pantcles at maturity diffuse, ellipsoid to ovoid,
4-21 cm long, 1-15 cmwide at the widest point, occasionally with glan-
dular pits on the rachis around one of the lower nodes, these rarely
also scattered along the rachis; primary branches capillary, straight
to drooping, at maturity spreading to ascending, rarely reflexed, the
lowermost or next above whorled or, infrequently, fascicled or paired,
or rarely both solitary, the upper alternate, opposite or whorled;
secondary branches usually present on the longer primary branches, at
maturity appressed to strongly spreading, straight to flexuous, rare-
ly bearing tertiary branches. Pedicels at maturity straight to flex-
uous, appressed to spreading. Spikelets narrowly ovate, the larger
3.5-10.0 mm long, 0.9-1.8 mm wide, consisting of 5-17 florets. First
glume 0.3-0.8 mm long, 1/4-1/3(-1/2) as long as the lowest lemma, the
second 0.7-1.2 mm long. JLemmas with inconspicuous lateral nerves,
usually pellucid, sometimes nearly opaque, with a pale to pronounced
grey-green cast and with the region near the apex red or purple, at
first erect to ascending, the dorsal side parallel to the longitudi-
nal axis of the spikelet to diverging 30° from it, at maturity more
divergent, the first or second 1.2-1.8 mm long. Paleas nearly al-
ways deciduous, sometimes tardily so, at maturity bent away from and
frequently exposing the rachilla. Caryopses smooth, yellow-brown to
24 Eragrostis pectinacea-ptlosa Complex
chestnut, pyriform to ovoid, slightly flattened laterally, 0.5-0.9mm
long. Pollendiameter 22.8-31.6u. Chromosome nurber: 2n=40. Platell.
Distribution and habitat: Weedy plants, growing mostly at elevations
below 1,000 m in disturbed, damp to wet, sandy soils from southern
Maine, southernmost Quebec, and the eastern side of Lake Ontario
south along the Appalachians and Atlantic Seaboard throughout the
West Indies, and west through the Gulf States to central Oklahoma
and eastern Texas. Also in the Isthmus of Tehuantepec region of
Mexico and in scattered locations elsewhere in the United States
and Mexico (Fig. 7).
ype: Scheuchzer, 1719, pl. IV, tab. 3. Seen.
In addition to its range in North and Central America, £. ptlosa
var. pilosa is found in the warm-temperate and tropical regions of
South America and the Old World. Since the present treatment is re-
stricted geographically to North and Central America, only names
which have been applied to American members of this taxon are con-
sidered in the synonymy, as it remains possible that study of the
Old World material of this variety will demonstrate the existence
of taxa to which some of these other names are applicable.
It is commonly stated that £. pilosa var. pilosa was introduced
into the New World in post-Columbian times. ‘This assumption is
based on the fact that many of our weeds were introduced from Eur-
asia, but in this case I can see no compelling reason to accept it.
Indeed, the fact that £. pilosa var. perplexa, a taxon endemic to
North America, is probably an evolutionary offshoot of F. pilosa
var. ptlosa is an argument against this assumption.
Harvey (1948), Hitchcock (1951), and Fernald (1950) recognized
E. pilosa sensu stricto and &. peregrina Wieg. (= E. multtcaults
Steud.) as related but distinct species. Harvey separated them on
the basis of the ratio of pedicel length to spikelet length, Hitch-
cock used spikelet width, and Fernald emphasized the presence or
absence of hairs on the leaf sheath auricles. Examination of speci-
Z5
losa var. pilosa
ts pt
Eragrost
[ on!
“Us9S Jou UsuUToeds {(gréT) AsATeH worzy SToquAs peut[iepug
*(setqetzea yj0q) vsoqid +g JO uoTINqTIzASTG
exeTdiod *iea esottd ack )
eso[td *1ea esottd -qo
“L
°3Ty
26 Eragrostts pecttnacea-pilosa Complex
mens of 'E. peregrina'' and FE. pilosa s.s. showed that within a sin-
gle plant, the pedicels of the lateral spikelets may vary from lon-
ger than to shorter than the spikelet; measurements of spikelet
width in the two entities showed no difference in means or ranges
(Table 4); and all plants on which glabrous auricles were found al-
so had some sheaths with long-haired auricles. There are, however,
differences between the two taxa. F. ptlosa s.s. is usually a tal-
ler plant with strongly divergent to loosely appressed pedicels,
while "E£. peregrina'' is shorter and has its pedicels and secondary
panicle branches appressed to the primary panicle branch. '"'F. per-
egrina'' is identical in all other respects to &. pilosa s.s., es-
pecially in having the primary panicle branches at one of the low-
est two nodes whorled, short glumes, deciduous paleas, the same
chromosome number, the same pollen diameter, and the same spikelet
morphology (see also Table 4).
In a uniform culture experiment comparing "'F. peregrina" and E.
pilosa s.s, caryopses from two collections of "F. peregrina," Koch
6664 (MICH) and Koch 6665 (MICH), produced mature plants which were
as tallas typical £. pilosas.s. and inwhich the secondary panicle
branches were spreading, but which retained the appressed pedicels of
their parents. The progeny of the £. pilosas.s. controls, Anderson and
Sternberg 3495 (MICH), Koch 65131 (MICH), and Koch 678 (MICH), were
typical. Thus, the only character separating "ZF. peregrina'' and E.
pilosa s.s which was not found to be environmentally controlled is
the angle of divergence of the pedicels. However, this character
is highly variable in £. pilosa s.s., as it is not uncommon to find
both appressed and spreading pedicels on the same plant.
These conclusions are further supported by the observation that
within a population of typical F. pilosa s.s., plants which are
growing under less favorable conditions usually have the appearance
of typical "EZ. peregrina," viz. small plants with their pedicels and
secondary panicle branches appressed to the primary branches. This
is especially common in plants that are subject to frequent crushing
because they are growing in a roadway or sidewalk.
Geographically, the £. peregrina phenotype is most commonly found
Eragrostts ptlosa var. ptlosa
TABLE 4
COMPARISON OF &. PILOSA S.S., "E. PEREGRINA," AND "E.
Character
Plant height (cm)
Panicle length (cm)
Panicle width at
widest point (cm)
Spikelet width at
anthesis (mm)
Number of florets
per spikelet
Spikelet length
(mm)
First glume length
(mm)
Second glume length
(mm)
Lemma length (mm)
Caryopsis length
(mm)
Embryo length as
fraction of cary-
opsis length
Pollen aiancter”
ea)
PERPLEXA"
WITH RESPECT TO QUANTITATIVE CHARACTERS*
Ew pilosa SsS. "BE. peregrina" "E. perplexa"
Mean Range Mean Range Mean Range
35:69 8-66 21.3 7-40 31.8 10-53
12.4 4-21 9.1 4-17 15.0 5-23
6.9 3-15 4.0 1-9 5.6 3-10
0.7 0.3-1.1 0.7 0.5-1.0 0.9 0.6-1.4
8.5 5-17 8.5 6-12 10.0 7-12
Del 3359=10.0 4.4 3.5-5.5 6.3 4.5-8.0
0.5 0.3-0.8 0.5 0.4-0.7 0.5 0.3-0.6
0.9 0.7-1.2 1.0 0.7-1.2 Le2 1.0-1.4
1.3 1.2-1.5 1.4 I-18 1.9 1.8-2.0
0.7 025-09 0.6 0.5-0.8 0.9 0.8-1.0
0.5 0.4-0.7 G.5 0.4-0.6 0.6 0.5-0.7
26.5 22.8-31.6 270 20.2527) <9 27 «2 25.5-29.2
(u)
a F
Sample sizes:
b
means "2. peregrina," and 5 sample means for
being 15 grains from a collection.
these sample means.
BE: pttosa (s.s.), 30; “E. peregrina,” 15;
"EH. perplexa,"
"Mean" is the mean of 30 sample means for F£. ptlosa (s.s.), 5 sample
"RE. perplexa,"
Likewise, "range" is the range of
each sample
28 Eragrostis pecttnacea-ptlosa Complex
along the east coast of the United States, from New England south
to Virginia, but collections from other localities are also known,
for example, Oregon [Suksdorf 2882 (US)], California [Hart s.n.,
July 1921 (US)], and Mississippi [Tracey 8386 (US)]. This pheno-
type is also found scattered throughout the Eurasian range of £.
pilosa S.S.
It appears that "FZ. peregrina’ is only an extreme of the varia-
tion found within £. ptlosa s.s. This form occurs most frequently
in the United States at the northern extreme of the range of £. pt-
losa sensu lato, perhaps in response to environment.
2. ERAGROSTIS PILOSA (L.) Beauv. var. PERPLEXA (Harvey) S. D. Koch
Eragrostts perplexa Harvey, 1954, p. 409.
Eragrostis pilosa (Linnaeus) Beauvois var. perplexa (Harvey) S. D.~
Koch combinatio nova. [Based on EF. perplexa Harvey.]
Differing from the typical variety in having glandular ptts
densely to sparsely scattered on panicle rachis and sometimes pri-
mary panicle branches, sheaths, at least on the midribs, abaxial
blade midribs, and usually on the culms; spikelets at anthesis wi-
der, 0.6-1.4 mm wide; second glumes longer, 1.0-1.4 mm long; lem-
mas longer, 1.8-2.0 mm long; caryopses longer, 0.8-1.0 mm long.
Plate Tit.
Distribution and habitat: Rare plants of the high plains of the
United States, growing near water in sandy or alkaline soils, most-
ly above 500 m altitude, from North Dakota south to southern Colo-
rado and northwestern Texas. Also one collection from the state of
Sonora, Mexico (Fig. 7).
Holotype: W. L. Tolstead s.n., in alkaline soil, Mellette County,
South Dakota, 30 August 1935. (US). Seen.
Neither viable seed nor living material of this variety was
available to the author, so the chromosome number could not be de-
Eragrostis ptlosa var. perplexa 29
termined. However, on the basis of the relationship between ploi-
dal level and pollen diameter illustrated in Fig. 6, a chromosome
number of 2n=40, the same as that of £. pilosa var. pilosa is ex-
pected.
Harvey (1948, p. 194), in describing this variety as a new spe-
cies, acknowledged its close affinities to #. pilosa: "Species £.
ptlosae affinis sed differt praecipue foveolis glandulosis numero-
sis in vagina, rachi ramisque paniculae.'' In examining the nine
available specimens belonging to this taxon, it was found that not
only does it have the diagnostic characters of F. pilosa, viz. pri-
mary panicle branches at one of the lowest two nodes whorled, deci-
duous paleas, short first glumes, small pollen diameter, and the
typical spikelet morphology, but also that the range of variation
of most of the other characters also falls within the range of the
typical variety of FE. pilosa (Table 4).
The major difference between the two varieties is that £. pilosa
var. perplexa has glandular pits scattered more or less over the
whole plant, especially on the midribs of the sheaths and blades.
Glands are also found on some specimens of £. ptlosa var. pilosa,
especially those from the Old World tropics, but these are always
restricted to the panicle axis or culm or both. There are also mi-
nor differences in the second glume length, first or second lenma
length, spikelet width at anthesis, and caryopsis length (Table 4).
Two specimens of £. pilosa var. perplexa are apparently inter-
mediate between the two varieties. These are Wiggins and Rollins
186 (US), from Sonora, Mexico, and Christ, Ward, and Frutchey 1679
(US), from Colorado. They have relatively few glands, those on the
sheaths and blades being restricted to a short segment of abaxial
midrib at the collar.
From Fig. 7, it appears that the range of £. ptlosa var. perplexa
is disjunct from that of the typical variety; the latter extends
west into eastern Texas, the former only barely gets into northwest-
ern Texas. However, the distances between the collections are so
great that the only meaningful statement that can be made about the
distribution of £. pilosa var. perplexa is that it is found in the
30 Eragrostts pectinacea-ptlosa Complex
Great Plains both within the United States and in their extension
into Mexico. It remains possible that the ranges of the two vari-
eties do overlap.
In summary, although there is a clear difference between the two
varieties in the distribution and abundance of glandular pits, the
high degree of morphological similarity in all other respects, cou-
pled with the presence of intermediates and the high probability of
overlapping distributions, argues for varietal rather than specific
status for Z. perplexa.
3. ERAGROSTIS PECTINACEA (Michaux) Nees
Poa pectinacea Michaux, 1803, p. 69.
Poa tenella sensu Nuttall, 1818, p. 67. [Non Linnaeus (1753). I
have not been able to locate a specimen which Nuttall examined.
Interpretation based on description. ]
Eragrostis brizoides Schultes, 1824, p. 319. [Based on Poa tenella
sensu Nuttall. ]
Poa nuttallit Kunth, 1829, p. 116. [Based on Poa tenella sensu Nut-
tall. Non Sprengel (1824, p. 344).]
Eragrostis purshit Hortus ex Schrader in Schlechtendal, 1838, p.
451. [Emendavit Gray, 1856, p. 564. Taken up by Gray when he
misapplied £. pecttnacea (Michaux) Nees to £. spectabilts (Pursh)
Steudel. |
Poa diandra Schrader in Schlechtendal, 1838, p. 451. [Pro synonymo
E. purshit Hortus ex Schrader. ]
Eragrostis nuttalltana Steudel, 1840, p. 563. [Based on Poa tenella
sensu Nuttall. ]
Eragrostis pectinacea (Michaux) Nees, 1841, p. 406.
Eragrostis pensylvanica Scheele, 1844, p. 58. [Type: Gardiner s.n.,
"in Pensylvania.'' Not seen. ]
Eragrostis peetinacea (Michaux) Steudel, 1854, p. 272.
Eragrostis cognata Steudel, 1854, p. 273. [Holotype: Frank s.n.,
collected in 1837 in Ohio. (P). Fragment seen at US.]
Eragrostis unionis Steudel, 1854, p. 273. (Holotype: Frank s.n.,
collected in 1835 at Miami, Ohio. (P). Fragment seen at US.]
Eragrostis pectinacea |
Eragrostts diffusa Buckley, 1862, p. 97. [Holotype: Buckley s.n.,
collected in Texas. (PH). Photograph and fragment seen at US.]
Eragrostis purshit var. delicatula Munro ex Lamson-Scribner, 1883,
p. 30. [Nomen nudum; £. diffusa Buckley cited as the only syno-
nym. |
Eragrostts purshit var. diffusa (Buckley) Vasey in Coulter, 1890,
P< - 29
Culms 10-60(-110) cm tall, vary rarely with a ring of glandular
pits beneath one or two of the nodes. Sheaths eglandular, mostly
shorter than the internodes, glabrous except for a sparse tuft of
white hairs at the apex of the margins. Blades glabrous, eglandu-
lar, the uppermost 2-15(-30) cm long, 1.0-4.5 mm wide at the widest
point. Panteles at maturity open, ovoid to pyramidal, sometimes
narrowly so, 5-25(-30) cm long, 3-12(-15) cm wide; primary branches
slender, straight, at maturity spreading to rarely ascending, rare-
ly somewhat reflexed, alternate or opposite, the lower rarely
whorled or fascicled; secondary branches usually present on longer
primary branches, fewer above, at maturity appressed or ascending,
rarely strongly divergent, in some cases bearing appressed to ascend-
ing tertiary branches. Pedicels at maturity appressed to the bran-
ches, rarely diverging as much as 20°. Sptkelets ovate to narrowly
so, the larger 4.5-11.0 mm long, 1.2-2.5 mm wide, consisting of 6-15
(-22) florets. First glwne 0.5-1.1 mm long, 1/2-3/4 as long as low-
est lemma, the second 1.1-1.7 mm long. Lemmas with moderately con-
spicuous lateral nerves, pellucid to opaque, grey-green or stramin-
eous, often with a purple or red cast, especially at the apex, at
first the dorsal side diverging approximately 30° from the longitu-
dinal axis of the spikelet, not changing with maturity, the lower
1.0-2.2 mm long. Pazeas persistent or, infrequently, tardily de-
ciduous. Caryopses smooth to inconspicuously longitudinally stri-
ate, yellow-brown to chestnut-brown, pyriform, slightly flattened
laterally, 0:5-1.1(-1.4) mm long. Pollen diameter 28.9-35.6u.
Chromosome number: 2n=60. Plate IV.
5Z Eragrostts pecttnacea-ptlosa Complex
Distribution and habitat: Weedy plants growing mostly at elevations
less than 2,500 m, in sandy or gravelly to loamy soils in disturbed
places from southern Maine west to eastern North Dakota and Nebraska,
central Colorado, southern Utah, and central and western California,
and south to Cuba, Puerto Rico, and Panama. Also occasional in oth-
er parts of the United States where it is introduced but apparently
does not maintain itself (Fig. 8).
Holotype: Michaux s.n., ''In arvis illinoensibus.'' (P). Fragment
seen at US.
In addition to the synonyms listed, Hitchcock (1935) included
Poa carolitntana Biehler (1807, p. 33) and its derivatives, EFragros-
tis carolintana (Biehler) Lamson-Scribner (Lamson-Scribner, 1894,
p. 49), and Eragrostis ptlosa var. carolintana (Biehler) Farwell
(Farwell, 1916, p. 182). However, as Fernald (1945) pointed out in
his paper announcing the discovery of Biehler's work, it is not at
all certain that this name applies to £. pectitnacea, for the de-
scription states that the spikelets have only five florets, while
those of £. peetinacea typically have 6-15 florets. It is possible
that Biehler had a specimen of £. frankiti, but this species rarely
has blades nine inches long, as the description requires. Chase
apparently agreed with Fernald, for in the 1950 edition of Hitch-
cock's Manual, these are included among the unidentified names.
Their final disposition will probably never be known, for it de-
pends upon the examination of the type, which was in Sprengel's
herbarium. Sprengel's grasses were at Berlin (B), and were de-
stroyed, although it is possible that the type is among the few
specimens in the Willdenow Herbarium, which is still extant, or is
in the small set at Lund (LD) (Stafleu, 1967, p. 455). In any case,
the name FE. peettnacea would not be affected.
Within £. peetinacea sensu lato, as I have circumscribed it, two
species have traditionally been recognized (e.g., Hitchcock 1951,
Harvey 1948). The first of these, F. pectinacea sensu stricto, is
basically an inhabitant of the eastern United States, while the
35
Eragrostis pectinacea
[ Yon!
—
'
\
°
lo
fe)
‘T°S vDaopui40ed +7 jo
uot§nqt4z3sTq
°8
34 Eragrostis pectitnacea-ptlosa Complex
other, "EZ. diffusa,'' occurs primarily in the southwestern United
States and Mexico. The two differ in size and in the density of
spikelets along the panicle branches, EF. pectinacea s.s. being smal-
ler and more delicate, and having the spikelets sparse along the
panicle branches, while "Z. diffusa'' is more robust and the spike-
lets overlap along the panicle branch (Figs. 9 and 10, left).
Examination of a large number of herbarium specimens showed that
(1) intermediates between the two species are abundant, and (2) it
is not uncommon to find members of "2. diffusa'' in the eastern Uni-
ted States and of F. pectitnacea s.s. in the Southwest and Mexico.
In addition, field observations demonstrated that, within a popula-
tion of either species, individuals growing in the more favorable
situations were taller and had less dense spikelets along the pan-
icle branches than those in less fortunate circumstances. In the
gypsum hills of south central Kansas, an area noted for the presence
of plants usually associated with the southwestern United States
(R. L. McGregor, University of Kansas, personal communication),
nearly all of the populations I observed included examples of both
species and intermediates.
Because of these observations, a uniform culture experiment was
performed in order to determine the extent of the influence of eco-
logical factors. Caryopses from 46 collections of Z. peetinacea
s.1., distributed over the whole range of the species and represent-
ing both extremes, as well as intermediates, were used. These are
listed in Appendix III of Koch (1969). In all cases, it was found
that the differences between the two traditional taxa became less
striking, although they did not altogether vanish. Under the fairly
moist, shady conditions in the greenhouse, extreme members of ''Z.
dtffusa'' maintained their larger size, but the density of spikelets
along the panicle branches decreased, while extreme examples of £.
pectinacea s.s. became taller, but the density of spikelets along
the panicle branches remained low (Figs. 9 and 10). It is clear
that the moderate conditions of greenhouse culture induced pheno-
typic changes which parallel those seen in the field in a single
population, viz. that in less exposed situations, extremely dense
“68GI Yoor
jo Auaso0id umox8-ssnoyueety :243TY * (6899 Yooy) WoTIDeTTOO PpTetd
:2je7T ‘owerqxe vsnffip +7, ‘qUewtiedxe aanqy[no wazostun °6 *3TA
a9
*R 7G *S 53H
"OSS9 NASH Wey “we “py THD
ageg ON OD Noy Cp mag “HD
Dg BF PRs ysedaq
ag “a t PICKS Ep
rea apiseq #22 4p
$959 Yooy 1802; HoIg
*iuesoy STUN ITP “3 UIP
saovip deep ‘Apweo CAat yes oy maMEZ ULL
“g¥a1 *HE tay0;9g
TQogT ME OeniyEDeIt By pees woz) uMRID
*seamy ‘toy pan ‘opt day fo
Kieu “im g°t ‘99 Voseuyiy 0% GOT HH “g'h; OAS
Os 48kP Sp IIe qoaKs erie TsoY Kner on ey
ras osrye ig Ame |
“STpTS Ty
SY prose Ig
SOON do LUSTRANG THA op Renee
ay
1 Ra HI SURETY
HOO CL MARLTLIS 40 SNOMLOTTIOS
Eragrostts pectinacea
VY
ee é
os
= &
:
* a €
b, *
4 3
e Axe 3
Eragrostts pectinacea-ptlosa Complex
‘OPITGS9 Yooy Jo Auas8oid
umois—ssnoyuseId :248TY *(OPLS9 YOOY) WOTIDPETTOO pTeTgq 33507
*eulaiqXe *S°S DaaDUIZOad *7 ‘jUewTredKxa |sanq{Nd. wArzostuq ‘OT °3Tg
sures "3K
DPISO MOA YOY “A *s *TLID Ob EQON “TED, IY Hg HO,
} prrigadag “pswA peoayre
te oP za ‘320080, eon “(oeTsy) Sapgoatod “| '(uETS9)
NVI “GRESS U2oe 14023 Prag WONIFIP °H (LETS) Pe tteseG “YO NET: ‘eenoyezeN
Tou were Apeye fdewn BE UDyseqIalod Tysmm avamvyy.
"2061 UF PHOe wOIz amioqus—erd uy wioarn
“soot “te 2aqe39q
Sy3s023vag
IRE *paed PROXT HRY FEF FINGeeTH UE polop
AVOUOH 40 AUSwAWwn JRL 40 KALAVEEH avo8 *Zagdhoy 109 azuepuadeRH pemayxn JHE mE)
* 9c aurpue In sates
OO
Ge eese—e a
|
| To yc apr? 40 ENT
| HOON CO NSHdUS 4O SNOHOTHOD
|
*
i
/
e
&
3
a
fie
if sd |
ey |
rit
/
4
4
4
«”*
~
o \
ss
‘ ‘
a”
”
‘ \
ae
‘
'
Eragrostts pectinacea 37
spikelets become less dense, and shorter plants taller.
The uniform culture experiment demonstrated that the most extreme
phenotypes in £. pectinacea s.l1. are at least in part environmental-
ly induced. However, it remains possible that there are two entities
within this species. In order to explore this, a modification of
Anderson's hybrid index method (Anderson, 1949) was used. Seven
characters, in which the extremes of the two traditional taxa differ
most, were selected. A random sample of 200 specimens was then ta-
ken, and the seven characters measured on each. The characters used
were:
1. . Plant height: height, in centimeters, of the tallest
mature culm.
2. Culm width: diameter, in millimeters, of the tallest
mature culm just below the lowest node.
3. Number of spikelets per centimeter of panicle branch:
number of spikelets on the longest panicle branch divided
by the length of the same branch in centimeters.
4, Spikelet length: length, in millimeters, of a single
large spikelet.
5. Lemma length: length, in millimeters, of the first or
second lemma on a large spikelet.
6. Caryopsis length: length, in millimeters, of a large
caryopsis.
7. Leaf width: width, in millimeters, of the uppermost
leaf of a mature culm, measured at its widest point.
These raw data were converted to summed standardized values by a
computer program, called HINDX, kindly written by Mr. G. William
Moore, using the University of Michigan's IBM 7090 computer. The
program directed the following operations:
1. For each character, the mean and standard deviation
were calculated.
2. The value of each character for each plant was then
converted to a "normal score" according to the equation,
normal score = —————— ;>
38 Eragrostts pectinacea-ptlosa Complex
where Xij is the value of the it character of the cis
plant and x; and s; are, respectively, the mean and stan-
dard deviation of the it® character for all plants. The
standard deviation, because it is a square root, can be
either positive or negative. Consequently, the sign of
the normal score is unspecified. The normal scores cor-
responding to those character states which are found in
extreme members of "E, diffusa" were assigned positive
signs, those found in extreme £. pectinacea s.s. were giv-
en negative signs. In addition, this step puts all char-
acters on the same scale by adjusting the mean of each
character to zero and the standard deviation to one. Thus,
the data were coded without the loss of information which
occurs when the range of variation is simply divided up
into segments, as in Anderson's original hybrid index.
h
3. The seven normal scores for each plant were summed
to give a summed standardized value. Because of the ad-
justment of the sign in the previous step, the normal
scores are additive, and the summed standardized value
is proportional to the position of the plant on a linear
scale between the most extreme £. pectinacea s.s. (lowest
value) and the most extreme "EZ, diffusa'' (highest value).
A frequency distribution of these summed standardized values was
then plotted (Fig. 11). It is clear that the distribution is nor-
mal and unimodal, although it is somewhat skewed to the right. This
shows that the variation within £. pectinacea s.1. is continuous and
fits the model of random variation about a single value, the mean.
For this reason, I conclude that £. pectinacea s.s. and "FE. diffusa"
are parts of a single taxon.
In support of this conclusion, it should be stated that I failed
to find any qualitative difference between the two traditional spe-
cies. In gross morphology, there is continuous variation in all the
characters I examined. In leaf epidermal anatomy and pollen diam-
eter, I found no difference. The chromosome number is 2n=60 for all
collections examined, regardless of phenotype.
Fig. 12 shows the geographic distribution of summed standardized
values. The data for this map were prepared as follows: First the
range of the species, sensu lato, was divided into squares, approx-
imately 300 miles on a side, except that somewhat larger squares were
used in areas where there are few collections or where irregularities
of the land mass or the range border caused the square to contain les
Ci ee Ae ey es
Eragrostts pecttnacea
Number of
individuals
© Ne] +t N jo)
(class midpoints)
Summed standardized value
Sample size
pectitnacea s.l.
ndardized values for #.
Frequency distribution of summed sta
Fig. ll.
200, mean
37).
See text for explanation (p.
0.69.
0.00, standard deviation =
-*(zy *d) ¢ eTqey wortz e}eq ‘uot JeUeTdxea Az0z (gE *d) 3xaq Vas
‘[T°S veovuizoed *7] UT UOTIETACA TRUTID “ZT *3Tgq
44
Cad O°T-
& onTea
’ poztpiepueys 0°0
: pewwns
“o O° T+
aa
TOUNAS JO NOLLVNV'IdXd
Eragrostis pectinacea-pilosa Complex
40
Eragrostts pecttnacea 4l
than 90,000 square miles. These were then numbered from north to
south and from east to west, giving 37 squares in all. All of the
specimens for which summed standardized values had been calculated
were then assigned to the appropriate squares, according to their
places of collection. Additional specimens, these also randomly
selected, were then measured to give each square a minimum of three
(except square 31, in which only two were available), and the data
were reprocessed by the HINDX program to obtain summed standardized
values for all of the specimens. The means and standards deviations
of the summed standardized values for each square were calculated
(Table 5). Bar graphs for each square were constructed, the length
of the filled portion being proportional to the mean summed stan-
dardized value.
In examining the map, it becomes clear that the geographic pat-
tern of variation in the mean summed standardized values is a cline.
Change takes place in two directions, one from northeast to south-
west in the United States and the other from northwest to southeast
in Mexico and Central America. Within the eastern deciduous forest
region of the United States, the mean values are all less than 0.00
and are relatively constant. In the Great Plains and Texas the val-
ues rise, reaching their peak in the southwestern United States and
the northern three-quarters of Mexico. In the southernmost square
of Mexico (square 26), all of Central America, and also Cuba, the
mean values decrease again, although they do not fall below 0.00.
Square 7, consisting for the most part of Florida, has an anomalous-
ly high value, considering that it is located in the eastern United
States. However, it should be noted that £. pectinacea s.1l. is rare
in all of Florida except the southernmost end. Most of the speci-
mens measured come from this region which is well known for its flor-
istic affinities to the Caribbean islands. The similarity in the
mean summed standardized values of squares 7 and 8 (Cuba) conforms
to this’ well-known affinity.
The distribution pattern of mean summed standardized values ex-
plains the apparent geographical difference between the two tradi-
tional taxa: they are the extremes in a pattern of clinal variation.
7-0
c9°0
78°0
88°0
09°0
€S°0
8L°0
99°0
Oe E
08°0
A oa |
ce*O
6e°T
c8°0
SS*0
00°T
L6°0
c6°0
Eragrostis pectinacea-ptlosa Complex
ucT jerTAep
paepueqs
TT*O
6T*O
80°O
OL°O
ce*0
€0°O
8L°0
S7"0
£9°0
£9°0
70°T
TT*O
78°0
TL°O
07°O
6S°0
O€°O
Z9°0
uray
LE
9€
GE
7%
ce
ce
TE
O€
67
82
Le
9¢
SC
7C
€?
CC
T?
0c
rn owyrtnnrennnrnrn szktOn gn MND YD tN SY
STENPTAT put aaen
JO zJequny
€S°O
€9°0
T9°O
97°0
c9°0
87°0
8L°0
£S°O
€9°0O
cv" 0
L£S°0
GS*0
LT°T
c8°0
£9°0
£9°0
c9°0
LE°O
09°0
bs uoT}eTAep
paepuejs
ZT *OId NO AUVADS HOVA UYOd
STVNGIAIGNI dO SYH@WAN GNV ‘SNOILVIAAG GYVGNVLS ‘SNVaW
¢ H1dVL
TS*O
ST*O
€7°0-
L£T°0O-
TC*0-
o7°O-
87°0-
Ge 0=
TS*0-
€
7
€C
L
7
€
9
7
S
80°0- 4
0
€
€
€
v7]
0
8
L
€
T
ee 0=
$0°0-
TO0°O
67° 0-
S9°0-
7€*O-
c7*0-
LGo°0=
ce. O=
c
T
uPoW STENPLTATput
JO Tequny
aN MO STN OO KW CO
oaenbs
Eragrostts pecttnacea 43
The lowest values are found in the eastern United States, the range
of traditional F. pectinacea. Similarly, the highest values are
found in the Southwest and northern Mexico, the area associated with
"EH. dtffusa.'' The values are intermediate in the central United
States and Central America, the areas of overlap in the distribu-
tions of the two traditional species.
In summary, five considerations argue against formally recogniz-
ing these two extremes as species, varieties, or subspecies: (1)
The uniform culture experiment showed that the marked differences
between the extremes are partially induced by environment. (2) The
frequency distribution of the summed standardized values is unimo-
dal, whereas some degree of bimodality would be expected if the
races were in fact differentiated enough to warrant formal recogni-
tion. (3) No known discontinuous character serves to separate them.
(4) Geographically, the two extremes grade into one another over two
large areas, the Great Plains and Central America. (5) The standard
deviations of the means in the squares are high and relatively uni-
form. Their magnitude is a reflection of the fact that it is not
unusual to find a member of one extreme in the area (and sometimes
even in the same population) primarily occupied by the other extreme.
The uniformity of the standard deviations is the result of the ab-
sence of a definite geographical zone of intergradation, a feature
which is further emphasized by the lack of any discernible geograph-
ic pattern in the distribution of unusually high and low standard
deviations.
With respect to the origin of this cline, two hypotheses present
themselves: (1) that two originally isolated entities have come in-
to contact, most probably through man's activities, and have under-
gone extensive introgression, increasing their variability to the
point of becoming a single, variable entity (viz. ''subspecies amal-
gamation,"' Payne, 1962, pp. 222-231); or (2) that we are dealing
with a single, widely distributed species which has undergone cli-
nal, geographical differentiation without the influence of man. I[
do not believe that the data available allow a choice to be made be-
tween these, but considering the rate at which man is modifying the
44 Eragrostts pectinacea-ptlosa Complex
environment to meet his own, relatively uniform needs, and consider-
ing the ease of transportation of the seeds of this group of weeds,
it is highly unlikely that differentiation will increase.
4, ERAGROSTIS TEPHROSANTHOS Schultes
Eragrostis tephrosanthos Schultes, 1824, p. 316.
Poa tephrosanthos Sprengel ex Schultes, 1824, p. 316. [Pro synon-
ymo £. tephrosanthos Schultes. ]
Poa polymorpha Sieber ex Schultes, 1824, p. 316. [Pro synonymo
E. tephrosanthos Schultes. ]
Eragrostts delicatula Trinius, 1836, p. 70.—Trinius, 1838, p. 73.
[Holotype: Rtedel s.n., ‘in cultis prope Rio Janeiro."' (LE).
Fragment of Riedel specimen from Trinius Herbarium seen at US.]
Poa havanensts Trinius ex Steudel, 1841, p. 360. [Nomen nudum.
Fragments of Hwnbolt 5298 from Willdenow Herbarium (B) annotated
"Poa havanensis" seen at US.]
Eragrostis parvula Steudel, 1854, p. 277. ([Holotype: Parrys 172.
(CN). Fragment seen at US. }
Eragrostis purshti Schrader var. genuina Fournier, 1886, p. 116.
[Pro parte. Based, in part, on £. tephrosanthos Schultes.]
Eragrostis purshit Schrader var. miserrtma Fournier, 1886, p. 116.
[Based on #. parvula Steudel. ]
Eragrostis purshtt Schrader var. pauciflora Fournier, 1886, p. 116.
[Nomen nudum. Holotype: Berlandier 43. (P). Isotype seen at
US: |
Eragrostis ptlosa (Linnaeus) Beauvois var. delteatula (Trinius)
Hackel in Stuckert, 1904, p. 133.
Eragrostis arida Hitchcock, 1933, p. 449. [Holotype: Httehcock
156650. (US)... Seen. |
Culms (6-)15-90 cm tall, infrequently with a ring of glandular
pits beneath a few of the nodes. Sheaths eglandular, mostly shorter
than their internodes, glabrous except for a sparse tuft of long
hairs at the apex of the margins. Blades glabrous, eglandular, the
uppermost 3-21 cm long, 1.0-5.0 mm wide at the widest point. Pan-
Eragrostts tephrosanthos 45
teles at maturity diffuse, ovoid to pyramidal, 4-25(-40) cm long,
2-18 cm wide, occasionally with a few glandular pits on the axis
beneath the lowest node; primary branches slender, straight, spread-
ing at maturity, alternate or opposite, rarely whorled; secondary
branches usually present on longer primary branches, spreading at
maturity, in some cases bearing spreading tertiary branches. Pedi-
cels at maturity straight to flexuous, ascending to divaricate.
Spikelets ovate, the larger 4.0-10.5 mm long, 1.1-2.3 mm wide, con-
sisting of 7-20 florets. First glwne 0.8-1.4 mm long, 1/2-3/4 as
long as the lowest lemma, the second 0.9-1.7 mm long. Lemmas with
moderately conspicuous lateral nerves, pellucid to opaque, grey-
green or stramineous, often with a purple cast especially at the
apex, at first the dorsal side diverging approximately 30° from the
longitudinal axis of the spikelet, not changing with maturity, the
lower 1.1-2.1 mm long. Paleas persistent. Caryopses smooth to in-
conspicuously striate, chestnut-brown, sometimes paler, pyriform,
somewhat flattened laterally, 0.6-1.1 mm long. Pollen diameter
27.1-33.8u. Chromosome number: 2n=60. Plate V.
Distribution and habitat. Weedy plants growing mostly below 1,500
m altitude, from Florida and the southern Gulf Coast west to north-
western Texas, New Mexico, and southeastern Arizona, south through-
out Mexico, Central America, and the West Indies. Also found spo-
radically farther north (Fig. 13).
Holotype: Sieber, Flora Martinique No. 33. (M). Seen. Isotypes:
WS: and.P., Seems
This species, as circumscribed here, includes two traditionally
recognized species, £. arida Hitchc., found primarily in the south-
western United States and Mexico, and £. tephrosanthos sensu stric-
to, which occurs in the West Indies, Central America, and at scat-
tered localities along the Gulf Coast of the United States. Hitch-
cock (1951) separates the two on the basis of spikelet shape (lin-
ear vs. ovate to linear); Harvey (1948) uses the ratio between spike-
losa Complex
tnacea-pt
Eragrostts pect
46
*soyzunsouydez *y JO uoTInqTIzISTG
“eT
Eragrogtts tephrosanthos 47
let length and pedicel length. In addition, F. tephrosanthos s.s.
is usually a smaller plant, with a more dense and lax panicle, and
smaller spikelets than "F. artda."
Examination of herbarium material showed that approximately a
third of the specimens are intermediate, and that the ''Z. artda"'
phenotype is relatively common in the West Indies and Central Amer-
ica. In addition, a search for characters which did not vary con-
tinuously between the taxa proved futile.
A uniform culture experiment was performed using seed of 11 dif-
ferent collections from the southwestern United States, Mexico, and
Central America. These collections represented both extremes and
intermediates (Table 6). In all cases, progeny of members of the
traditional "#. arida'' were essentially identical to their parents.
The progeny of &. tephrosanthos s.s. and the intermediates, however,
were more robust than their parents, and had the more open panicle
with stiffly spreading pedicels and branches of "F. arida."
It appears then, that the two traditional species are not dis-
tinct taxa, but that F. tephrosanthos s.s. is merely a depauperate
form of "EZ. arida."’ This conclusion is supported by the observation
that seed production in many specimens of £. tephrosanthos s.s. 1S
low, while "FZ. artda'' and the greenhouse-grown progeny of £. teph-
rosanthos s.S. produce abundant seed.
The apparent range difference between the two traditional species
can be explained by assuming that £. tephrosanthos s.1. is adapted
to the relatively dry conditions of the southwestern United States
and Mexico. Here the &. tephrosanthos s.s. phenotype is rare, prob-
ably because the environment is relatively uniform over wide areas.
In the West Indies, on the Gulf Coast of the United States, and in
Central America, the whole range of variation is present. In these
places, the climate varies from wet to dry, and often does so over
short distances, depending upon such factors as the position of
mountains and direction of prevailing winds. Here the chance of
seeds being distributed into climatically suboptimal areas is great,
with the result that depauperate specimens are common.
As can be seen from the descriptions of £. tephrosanthos and £.
Ag Eragrostts pectinacea-ptlosa Complex
TABLE 6
SOURCE OF 2. TEPHROSANTHOS S.L. SEED USED IN
THE UNIFORM CULTURE EXPERIMENT
Collector and number;
Location Phenotype herbarium
California: Merced Co. "arida" Koch 69127; NSC
Costa Rica: Prov. Guanacaste intermediate Koch 6948; NSC
Costa Rica: Prov. Limon intermediate Koch 6919; NSC
Costa Rica: Prov. Limon tephrosanthos Koch 6923; NSC
Costa Rica: Prov. Puntarenas intermediate Koch 69673; NSC
Costa Rica: Prov. Puntarenas tephrosanthos Koch 6972; NSC
Costa Rica: Prov. Puntarenas tephrosanthos Koch 6974; NSC
Honduras: Depto. Tela
Atlantica tephrosanthos Koch 6984; NSC
Honduras: Depto. Tela
Atlantica “artda" Koch 6992; NSC
Mexico: Nayarit intermediate McVaugh 19135; MICH
Mexico: Sonora "arida" Gould 12050; TAES
“Phenotype of parent plant. "Intermediate" = between "Z#. arida"
and £. tephrosanthos s.s.
pectinacea, the only difference between these species is that the
pedicels of the spikelets which are borne laterally on the panicle
branches are appressed in £. pectinacea, while they are spreading to
various degrees in £. tephrosanthos s.1. This character in &. teph-
rosanthos S.l1. is expressed only at maturity, and even then—although
infrequently—most of the pedicels on a plant can be appressed. But
even in these cases, there is usually a panicle or panicle branch
somewhere on the plant which has strongly spreading pedicels. In &.
Eragrostis tephrosanthos 49
pectinacea, however, strongly spreading pedicels are never found
on any part of the plant.
This problem is relevant to the typification of £. tephrosan-
thos s.1., since two of the three plants on the herbarium sheet
which is the holotype and all of the isotypes which I have examined
have pedicels that are appressed to the panicle branch, the charac-
ter state typical of F. pectitnacea. Fortunately, one of the three
plants (the uppermost) on the holotypic sheet has spreading pedi-
cels, allowing the name to be unambiguously assigned to this spe-
cies. There are other characters which can be used to assign these
atypical specimens to F. tephrosanthos s.1.; namely, the spikelets
are generally smaller in West Indian £. tephrosanthos s.1. than in
West Indian £. pectinacea, and the number of spikelets per unit
length of panicle branch is somewhat greater in £. tephrosanthos
s.l., but these characters are much more ambiguous than pedicel
angle.
Although the distributions of the two species are coincident
over most of the southern part of the range of #. pectinacea (Figs.
8 and 13), and their chromosome numbers, flowering times, and hab-
itats are the same, no evidence of hybridization was found. Where
populations of the two species were found growing completely inter-
mixed, intermediates, although sought, were not found; all of the
plants had either spreading or appressed pedicels.
Progeny tests also failed to provide conclusive evidence of hy-
bridization. Table 7 shows that all 23 plants collected in Cali-
fornia from the areas of contact of two different mixed populations
of FE. pectinacea and the "'Z. arida'' phenotype of £. tephrosanthos
essentially bred true. None of the progeny had weakly spreading
pedicels as would be expected of hybrids if pedicel angle were un-
der the control of a polygenic or multiple allele system. Similar-
ly, only one of the 179 progeny of &. tephrosanthos had appressed
pedicels, and only two of the 240 progeny of £. pectinacea had
spreading pedicels. The last result would be expected if hybridi-
zation were occurring and pedicel angle were under the control of
a Single-locus, two-allele system.
50 Eragrostts pectinacea-ptlosa Complex
TABLE 7
PROGENY TEST OF PLANTS FROM MIXED POPULATIONS OF
EB. PECTINACEA AND E. TEPHROSANTHOS
Parents Progeny
Name, collection Sample Pedicel Number with Number with
number, and herbarium size angle spreading appressed
pedicels pedicels
Population I
E. peettnacea; Koch
69126 (NSC) 8 appressed 2 152
E. tephrosanthos;
Koch 69127 (NSC) 6 spreading 103 0
E. tephrosanthos;
Koch 69128 (NSC) 1 spreading 20 0
Population II
E. peettnacea; Koch
69131 (NSC) 5 appressed 0 86
E. tephrosanthos;
Koch 69132 (NSC) 3 spreading 55 1
Little significance is attached to the three plants that were dif-
ferent from their parents, primarily because none of the 416 other
progeny plants shared this pattern. They could well be the result of
accidental mishandling of the very small seeds during planting. Even
if they are hybrids, gene flow between the two species cannot be very
great, considering that they are rare, and that the plants selected
as parents in this test were those which had the greatest chance of
hybridizing. In addition, progeny of other collections of £. pectina-
cea and E. tephrosanthos, which were grown at various times, never
showed segregation for pedicel angle.
Their apparent genetic isolation argues forcefully, I feel, for
the recognition of &. tephrosanthos as a species distinct from F. pec-
Eragrostis lutescens Dull
tinacea. It is difficult to explain the absence of extensive in-
trogression in the broad area of overlap of the two ranges if the
two are not genetically isolated. The morphological difference be-
tween EF. tephrosanthos and £. pectitnacea is a relatively minor one
—spreading versus appressed pedicels—but the difference is con-
sistent so that there is little difficulty in separating the two.
Furthermore, there is a distributional difference which implies a
physiological divergence. This, quite possibly, is more signifi-
cant in terms of the evolutionary success of the two species than
is the morphological one, and, as such, should be given weight in
considering their taxonomic status.
5. ERAGROSTIS LUTESCENS Lamson-Scribner
Eragrostts lutescens Lamson-Scribner, 1899, p. 7.
Culms 6-25 cm tall, moderately to densely covered with glandu-
lar pits. Sheaths mostly less than the length of their internodes,
glabrous or with a sparse tuft of white hairs at the apex of the
margins, the nerves with many glandular pits. Blades glabrous, the
uppermost 2-12 cm long, 1.0-3.0 mm wide at the widest point, abaxial
surface and margins—especially toward the base—moderately to
densely covered with glandular pits, those on the margins raised
only very slightly, if at all. Panicles dense, narrowly ellipsoid,
at maturity 5-13 cm long, 0.5-2.0 cm wide; primary branches usually
alternate, at maturity ascending to appressed to the rachis; secon-
dary branches few, at maturity appressed to the primary branch; ter-
tiary branches usually lacking; axis and branches with few to many
scattered glandular pits. Pedicels at maturity appressed to bran-
ches to ascending, with a few glandular pits. Sptkelets narrowly
ovoid, the larger 4.5-7.5 mm long, 1.5-2.0 mm wide, consisting of
8-14 florets. First glume 0.9-1.4 mm long, 1/2-3/4 as long as the
lowest lemma, the second 1.5-1.8 mm long. Lemmas with moderately
conspicuous lateral nerves, pellucid, stramineous, often with a
purple cast, the lowest 1.7-2.0 mm long. Paleas persistent. Cary-
opses smooth, chestnut-brown, pyriform, slightly flattened lateral-
52 Eragrostis pectinacea-ptlosa Complex
ly, ca. 0.8 mm long. Pollen diameter 31.0-35.8u. Plate VI.
Distribution and habitat: Rare plants of sandy soils, mostly along
rivers of the Intermontane Basin from Washington and Idaho south to
Nevada and Arizona (Fig. 14).
Holotype: Piper 2624, Snake River, Almota, Washington, 7 July 1897.
(US). Seen.
6. ERAGROSTIS FRANKII Meyer ex Steudel
Poa parviflora sensu Nuttall, 1818, p. 67. [Non Pursh (1814, p. 80),
nec R. Brown (1810, p. 180). Specimen annotated by Nuttall ''Poa
parviflora,'' from his herbarium (BM), seen. ]
Poa micrantha Schultes, 1824, p. 305. [Non Eragrostis micrantha
Hackel (1895, p. 389). Based on Poa parviflora sensu Nuttall.]
Evagrostis frankit Meyer ex Steudel, 1854, p. 273.
Eragrostts erythrogona Nees in Steudel, 1854, p. 273. [Holotype:
Drummond s.n., St. Louis. ''Poa A."' (B). Fragment seen at US.
Reduced to synonymy under £. frankti Meyer ex Steudel by Gray
(1856, p. 564).]
Eragrostts captllaris (Linnaeus) Nees var. frankizi (Steudel) Far-
well, 19lo,; Pp. Lez.
Eragrostis frankii var. brevtpes Fassett, 1932, p. 95. [Holotype:
Fassett 12899. (WIS). Isotype: (US) Seen. ]
Culms 10-50 cm tall, usually with a few glandular pits beneath
some of the nodes. Sheaths mostly longer than their internodes, on
most plants at least a few with glandular pits along the midrib,
usually with a tuft of white hairs at the apex of and sometimes ex-
tending down the margins, seldom entirely glabrous. Blades glabrous
or infrequently sparsely pilose near the base of the adaxial surface,
the uppermost 2-10(-21) cm long, 1.0-4.0 mm wide at the widest point,
sometimes with a few glandular pits in the abaxial surface of the
midrib near the base. Pantcles at maturity ellipsoid to narrowly so,
open, compactly branched, 4-20 cm long, 2-9(-14) cm wide at the
5S
Eragrostts franktt
‘uses Jou uowtToeds {(gr6T) APATeH WOT STOqUAS poeUTTIepuy *Ssueoseqn] “WY
‘yT °3Ta
pue ‘ta7bu1ad *q Sityupaf “gq Jo suoT INnqTzIsTq
TeTsutad ‘ag
a
U
i
4 “
2 i. sus0SeINT “ge
é
| 7
*. 4 ¥ v ee el a
»S = oO
‘ ° care
Sec
A
i]
os
‘A
4
‘ en
Sale *
BS ;
ant ea a ce \
\ : _
1 ! { '
\ ) ----4 ;
! Oir~e- '
yoy
1 I !
ae ee ae o; ed
Se ce 7]
o! poe
:
' t
Oo, '
oe :
ee
H
Me
he eas 7
1 !
o |
‘ SS
pa ~-W- ~--4
\ '
t
1 H
TINGE nae iss
1
' '
1 '
; wo
, t
7
54 Eragrostts pectinacea-ptlosa Complex
widest point, sometimes with a few glandular pits on the axis be-
neath the nodes and scattered along the branches; primary branches
capillary, the lowermost and next above solitary, rarely paired,
the upper alternate, opposite, or whorled, spreading at maturity;
secondary branches usually present on the longer primary branches,
spreading at maturity; tertiary branches lacking. Pedicels spread-
ing at maturity, usually conspicuously exceeding the length of the
spikelet, usually with one or two glandular pits. Spikelets broadly
ovate, the larger 2.0-3.5(-5.6) mm long, 1.0-2.5 mm wide, consisting
of 3-6(-9) florets. First glwne 1.0-1.5 mm long, 3/4-1 1/4 as long
as the lowest lemma, the second 1.0-1.8 mm long. Lemmas with mod-
erately conspicuous lateral nerves, usually nearly opaque, grey-
green, often with a magenta cast, especially near the apex, the
lowest 1.1-1.6 mm long. Paleas persistent. Caryopses chestnut-
brown, dull, with minute longitudinal striae, broadly ellipsoid to
nearly globose, flattened laterally and dorsally, the dorsal side
frequently shallowly indented, 0.4-0.6 mm long. Pollen diameter
22.6-28.1(-29.5)u. Chromosome number: 2n=40, rarely 80. Plate
VII.
Distribution and habitat: In wetter areas, around streams and lakes
from Vermont to southern Minnesota, south to Virginia, northern
Georgia, and southern Arkansas (Fig. 14).
Holotype: Frank s.n., Ohio, collected in 1835. (P). Fragment seen
at US. Isotype: BM. Seen.
The usual chromosome number in £. frankii is 2n=40, the tetraploid
level in the genus. However, in one population from central Missouri
[Koch 6566 (MICH) pro parte], octoploids, 2n=80, were found along
with the usual tetraploids. Chromosome pairing in the octoploids was
apparently normal, 92% of the pollen was both stainable with cotton
blue in lactophenol and normal morphologically, and abundant viable
seeds were produced. Phenotypically, the octoploids differ from
tetraploid FE. frankii by their more robust habit, by the larger
Eragrostts pringlet oD
spikelets on a plant bearing 5-8 florets instead of the typical 3-6,
by having caryopses which are on the broadly elliptic extreme of the
normal variation from globose to broadly elliptic, by having panicles
that are less dense than typical because of a lower number of spike-
lets per unit length of panicle branch, and by a mean pollen diameter
of 29.5u instead of 22.6-28.1ly of typical F. frankit. However, there
is no discontinuity between the tetraploids and octoploids in any of
these characters.
From the description of £. frankit var. brevipes Fassett, it would
seem that this octoploid population is what Fassett was describing,
since this variety is said to differ from the typical one in having
longer spikelets. However, examination of an isotype and paratype
of £. frankit var. brevipes in the United States National Herbarium
showed that Fassett's variety is not the same as these octoploids.
The type material had panicles which were denser than typical £,
frankit, not more open—as is the case in the octoploids—and, al-
though a few of the spikelets are longer than typical, most fall
within the normal range of typical Ff. frankizi. But in the octoploids
the majority of the spikelets are unusually large. In addition, the
specimens of &. frankiti var. brevipes were not noticeably more robust
than typical £. frankit, while the octoploids are.
The apparent regularity of meiosis indicated by the high percen-
tage of normal pollen and the lack of any unusual chromosome pairing
behavior suggests that the octoploids arose by allopolyploidy involv-
ing a cross between £. frankii and some other species. A morpholog-
ically suitable second parent proved impossible to find, however,
because in all critical characters the octoploids are identical to
tetraploid £. frankti. Consequently they are believed to be auto-
polyploids.
7. ERAGROSTIS PRINGLEI Mattei in Mattei and Tropea
Eragrostis pusilla Lamson-Scribner ex Beal, 1896 [November], p. 481.
[Non Hackel (1896 [September], p. 27). ]|
Eragrostis pringlet Mattei in Mattei and Tropea, 1909, p. 241.
[Based on Eragrostis pusilla Lamson-Scribner ex Beal. ]
56 Eragrostis pectinacea-ptlosa Complex
Eragrostts sertbneriana Hitchcock, 1913, p. 361. [Based on Eragros-
tts pustlla Lamson-Scribner ex Beal. ]
Culms 6-26 cm tall, eglandular. Sheaths longer than their inter-
nodes, densely papillose-hirsute, eglandular. Blades, especially
near the margins, densely invested with long, pustulate hairs, eg-
landular, the uppermost 1.5-4.5 cm long, 1.0-4.0 mm wide at the
widest point. Panicles at maturity narrowly ellipsoid, open, com-
pactly branched, 3.5-10.0 cm long, 0.5-2.5 cm wide at the widest
point; primary branches capillary, the lowest and next above soli-
tary, rarely paired, sometimes distant, the upper alternate, oppo-
site, or whorled, spreading at maturity; secondary branches usually
lacking. Pedicels spreading at maturity, exceeding the length of
the spikelet, with a glandular ring near the midpoint. Spikelets
broadly ovate, the larger 1.7-2.3 mm long, 1.3-2.0 mm wide, con-
Sisting of 3-5 florets. First glume 0.8-0.9 mm long, 1/2-3/4 as
long as the lowest lemma, the second 0.8-1.1 mm long. Lemmas with
inconspicuous nerves, opaque at maturity, pale green to brown-red,
the lowest 1.0-1.4 mm long. Paleas persistent. Caryopses chestnut-
brown, whitened, with minute longitudinal striae, globose, somewhat
laterally compressed, 0.5 mm long. Pollen diameter 20.6-26.1n.
Place Vili.
Distribution and habitat: Known only from Guadalajara, Jalisco,
and Morelia, Michoacan, both in central Mexico (Fig. 14, p. 53).
Holotype: Pringle 2327, dry, rocky hills near Guadalajara, Mexico,
28 October 1889. (MSC). Seen.
CITATION OF SPECIMENS
Appendix II in the unabridged version of this paper (Koch, 1969,
p. 152-180) is a list of specimens representative of each taxon in
the £. pecttnacea-ptlosa complex. A limited number of copies of
this appendix are available from the author upon request.
Literature Cited 57
LITERATURE CITED
Anderson, E. 1949. Introgressive hybridization. New York: Wiley
and Sons. 109 p.
Beal, W. J. 1896. Grasses of North America. Vol. 2. New York:
H. Holt and Co. 706 p.
Beauvois, P. de. 1812. Essai d'une nouvelle Agrostographie. .. .
Paris. 182 p.
Bentham, G., and J. D. Hooker. 1883. Genera plantarum. Vol. 3.
London: L. Reeve and Co. 1,258 p.
Biehler, J. F. T. 1807. Plantarum novarum ex herbario Sprengelii
centuriam. . . . As "Novarum plantarum ex herbario meo centuria,"
p. 27-58. In K. P. J. Sprengel, Mantissa prima florae halensis.
Halle. 58+31 p.
Boissier, E. 1884. Flora orientalis sive. ... Vol. 5. Geneva:
Georg. 868 p.
Bonnet, Dr. 188]. Note sur l'Fragrostis pilosa P. B. variété da-
mienstana. Le Naturaliste 3:412.
Bowden, W. M., and H. A. Senn. 1962. Chromosome numbers in twenty-
eight grass genera from South America. Can. J. Bot. 40:1115-1124.
Brown, R. 1810. Prodromus florae Novae Hollandiae et Insulae van-
Diemen, . . . Vol. 1. London: Johnson. 592 p.
Buckley, S. B. 1862. Descriptions of plants. No. 3. Gramineae.
Proc. Acad. Nat. Sci. Phila. 1862:88-100.
Coulter, J. M. 1890. Upon a collection of plants made by Mr. G. C.
Nealley, in the region of the Rio Grande, in Texas, from Brazos
Santiago to El Paso County. Contrib. U.S. Natl. Herb. 1:29-61.
Farwell, O. A. 1916. Contributions to the botany of Michigan No.
14. Rept. Mich. Acad. Sci. 17:167-182.
Fassett, N. C. 1932. Notes from the herbarium of the University
of Wisconsin—VIII. Rhodora 34:95-96.
Fernald, M. L. 1945. Some inconvenient upheavals of familiar names
and author citations. Rhodora 47:197-204.
58 Eragrostis pectinacea-ptlosa Complex
1950. Gray's manual of botany. 8th ed. New York: Amer-
ican Book Co. Ilxivt+1l632 p.
Fournier, E. 1886. Mexicanas plantas. Pars secunda. Gramineae.
Paris. 160 p.
Gould, F. W. 1958. Chromosome numbers in southwestern grasses.
Am. J. Bot. 45:757-767.
1965. Chromosome counts in Mexican grasses. Bol. Soc.
Botan. Méx. 29:49-62.
Gray, A. 1848. Manual of botany of the Northern United States.
Boston and Cambridge: James Munroe and Co. /710 p.
1856. Manual of botany of the Northern United States.
2nd ed. New York: G. P. Putnam and Co. 739 p.
Hackel, E. 1895. Gramineae, p. 377-393. In H. Schinz. Beitrage
zur Kenntnis der afrikanischen Flora. Bull. Herbier Bois. 3:373-
441.
1896. Gramineae, p. 10-28. Jn H. Schinz. Die Pflanzen-
welt Deutsch-Siidwest Afrika. Bull. Herbier Bois. 4, append.
3:1-5/.
1902 ["1901"] Nr. 115. Eragrostis pilosa P.B. Agrost, p.
162 (1812) var. condensata Hackel nov. var. Allgem. botan. Z.
7213-
Harvey, L. H. 1948. Eragrostis in North and Middle America. Un-
published Ph.D. dissertation, Univ. of Michigan. Ann Arbor,
Mich. Univ. Microfilms. 2/0 p.
1954. New entities in North and Middle America Eragrostts
(Gramineae). Bull. Torrey Botan. Club 81:405-410.
Hitchcock, A. S. 1905. The identification of Walter's grasses.
Ann. Rept. Missouri Botan. Garden 16:31-56.
1909. The grasses of Cuba. Contrib. U.S. Natl. Herb. 12:
183-258.
1913. Mexican grasses in the United States National Her-
barium. Contrib. U.S. Natl. Herb. 17:181-389.
1930. The grasses of Central America. Contrib. U.S. Natl.
Herb. 24:557-762.
. 1932. The grasses of the Muhlenberg herbarium. Bartonia
L143 27-69.
Literature Cited 59
a 1933. New species and new names of grasses from Texas.
J. Wash. Acad. Sci. 23:449-456.
- 1935. Manual of the grasses of the United States. U.S.
Dept. Agr. Misc. Publ. 200. 1,040 p.
. 1936. Manual of the grasses of the West Indies. U.S.
Dept. Agr. Misc. Publ. 243. 439 p.
- 1951 ["1950"]. Manual of the grasses of the United States.
2nd ed. Revised by Agnes Chase. U.S. Dept. Agr. Misc. Publ.
200. 1,051 p.
Koch, S. D. 1969. The Eragrostis pectinacea-ptlosa complex in
North and Central America. Ph.D. dissertation, Univ. of Michi-
gan. (L. C. Card No. Mic 69-18042.) 191 p. Amn Arbor, Mich.:
Univ. Microfilms.
Kunth, C. S. 1829. Voyage aux régions équinoctiales du Nouveau
Continent, fait en 1/99-1804. Partie 6. Botanique, section 6.
Revision des Graminées. Paris. 5/79 p.
Lamson-Scribner, F. 1883. A list of grasses collected by Mr. C.
G. Pringle in Arizona and California, with descriptions of those
species not already described in American publications. Bull.
Torrey Botan. Club 10:29-32.
1894. Gramineae. Jn "List of Pteridophyta and Spermato-
phyta growing without cultivation in northeastern North America.
Mem. Torrey Botan. Club 5:27-59.
2 1899. New species of North American grasses. U.S. Dept.
Agr., Div. Agrostology Circ. 9:1~/7.
Lanjouw, J., and F. A. Stafleu. 1964. Index herbariorum. 5th ed.
Utrecht: Keimink en Zoon N. V. 251 p.
Link, J. H. F. 1827. Hortus regius botanicus berolinensis. Vol.
1. Berlin. 384 p.
Linnaeus, C. 1753. Species plantarum. 2 v. Holmiae: Laurentii
Salvii.
Mattei, G. E., and C. Tropea. 1909. Richerche e studi sul genere
Eragrostis, in rapporto ai nettarii estranuzali. Reale Istituto
Botanico di Palermo. Contrib. Biol. Vegetale 4:205-284.
Metcalfe, C. R. 1960. Anatomy of the monocotyledons. Vol. 1.
Gramineae. Oxford: Clarendon Press. 731 p.
Michaux, A. 1803. Flora boreali-americana. .. . Vol. 1. Paris:
Levrault. 330 p.
60 Eragrostts pectinacea-ptlosa Complex
Muhlenberg, H. 1817. Descriptio uberior graminum. . . . Philadel-
phia: Solomon W. Conrad. 295 p.
Nees von Esenbeck, C. G. D. 1841. Flora Africae australioris il-
lustrationes monographicae. Vol. 1. Gramineae. Glogau. 390 p.
Nicora, E. G. 1941. Glandulas epidermicas de algunas especies de
Eragrostts. Darwinia 5:316-321.
Nuttall, T. 1818. The genera of North American plants. ... Vol.
1. Philadelphia: De Heartt. 312 p.
Ono, H. and T. Tatecka. 1953. Karyotaxonomy in the Poaceae. lI.
Chromosomes and taxonomic relationships in some Japanese grasses,
Botanical Magazine, Tokyo 66:18-27.
Payne, W. W. 1962. Biosystematic studies of four widespread weedy
species of ragweeds (Ambrosia: Compositae). Ph.D. dissertation,
Univ. of Michigan. (L. C. Card No. Mic 63-423.) 331 p. Ann
Arbor, Mich.: Univ. Microfilms.
Pilger, R. 1956. Die natiirlichen Pflanzenfamilien. Band 14d.
Zweite Auflage. Berlin: Duncker and Humbolt. 225 p.
Prat, H. 1936. La systématique des graminées. Ann. Sci. Nat.,
10e serie, Botanique 18:165-258.
Pursh, F. 1814. Flora americae septentrionalis. 2 v. London:
White, Cochrane and Co.
Reeder, J. R. 1957. The embryo in grass systematics. Am. J. Bot.
44:756-/768.
1971. Notes on Mexican grasses IX. Miscellaneous chromo-
some numbers—3. Brittonia 23:105-117.
Scheele, A. 1844. Beitrdge zur Kenntniss der Grazer, vorztiglich
der deutschen und schweizerischen Flora. Flora 27:49-64.
Scheuchzer, J. 1719. Agrostographia, sive graminum, juncorum. .
Historia. Zurich: Tiguri. 512 p.
Schlechtendal, D. F. L. von. 1838. Reliquiae Schraderianae. Lin-
naea 12:353-476.
Schultes, J. A. 1824. Mantissa in volumen secundum systematis
vegetabilium Caroli a Linné. Stuttgart: Cotta. 388 p.
Sprengel, K. P. J. 1824 ["1825"]. Caroli Linnaei. . . . Systema
vegetabilium. Editio decima sexta. Vol. 1. Gottingen. 992 p.
Literature Cited 61
Stafleu, F. A. 1967. Taxonomic literature. Utrecht: Interna-
tional Bureau for Plant Taxonomy and Nomenclature. 556 p.
Steudel, E. G. 1840-1841. Nomenclator botanicus. .. . 2nd ed.
2v. Stuttgart, Tiibingen. [Vol. 1 published in 1840, Vol. 2
in late 1840 and 1841 (Stafleu, 1967, p. 458). ]
- 41854. Synopsis plantarum glumacearum. Vol. 1. Stuttgart:
J. B. Metzler. 475 p.
Stuckert, T. 1904 ["1905"]. Contribucién al conocimiento de las
Graminaceas Argentinas. An. Mus. nac. B. Aires, serie 3, 4:43-
161.
Tateoka, T. 1965. Chromosome numbers of some east African grasses.
Am. J. Bot. 52:864-869.
Thellung, A. 1907. Beitrage zur Kenntnis der Schweizerflora (VIII).
1. Beitrage zur Adventivflora der Schweiz. Viert. naturf. Ges.
Ziirich 52:434-473.
__+ 1928. Eragrostis Damienstana Ed. Bonnet. Repert. Spec.
Nov. Regn. veg. 24:323-332.
Trinius, C. B. 1836. Extraits de mémoirs. 15. Graminum in actis
academicis a se editorum generibus et speciebus supplementa.
Bulletin Scientifique, série I, publie par 1'Académie Impériale
des Sciences de Saint Pétersbourg 1 (9):65-71.
. 1838. Graminum in hisce actis a se editorum generibus ac
speciebus supplementa. Mém. Acad. Imp. Sci. Saint Pétersbourg,
série 6, 4 (2):1-107.
Wiegand, K. M. 1917. A new species of Eragrostis of the Old World
and North America. Bull. Torrey Botan. Club 19:93-96.
APPENDIX
63
List of Collections from Which Chromosome
Herbarium refers to the location of the voucher.
Counts Were Obtained
Entries followed
by a reference in parentheses have been taken from the literature;
vouchers of these have not been seen unless followed by !.
somatic chromosome number.
Taxon
E. frankit
E. peetinacea
2n
40
40
40
40
40
40,80
60
60
60
Locality
ILLINOIS:
Lake Co.
ILLINOIS:
Logan Co.
MICHIGAN:
Monroe Co.
MICHIGAN:
Monroe Co.
MICHIGAN:
Monroe Co.
MISSOURL:
Montgomery
Co.
ARIZONA:
Apache Co.
ARKANSAS :
Independence
Co.
CALIFORNIA:
Mendocino Co.
2n is
Collector & number;
herbarium; (reference)
Koch 675; MICH
Koch 6552; MICH
Koch 65145; MICH
Koch 6668; MICH
Koch 6670; MICH
Koch 6566; MICH
Demaree 38474; US
Koch 65140; MICH
Raven & Snow 136783; US
64 Eragrostis pecttnacea-ptlosa Complex
Taxon 2n Locality Collector & number;
herbarium; (reference)
E. peettnacea 60 FLORIDA: Koch 66333 MICH
(cont'd) Dade Co.
60 ILLINOIS: Koch 6550; MICH
Logan Co.
60 KANSAS: Koch 65753; MICH
Barber Co.
60 KANSAS: Koch 6589; MICH
Comanche Co.
60 KANSAS: Koch 6573; MICH
Douglas Co.
60 KANSAS: Barker 18773; KANU
Morton Co.
60 LOUISIANA: Koch 65129; MICH
Allen Parish
60 MASSACHUSETTS Koch 6666; MICH
Suffolk Co.
60 MICHIGAN: Koch 66/71; MICH
Monroe Co.
60 MICHIGAN: Koch 6667; MICH
Washtenaw Co.
60 MISSOURI: St. Koch 6561; MICH
Charles Co.
60 NORTH CAROLINA: Koch 6/712; MICH
Durham Co.
60 TEXAS: Palo Gould 7600; TAES
Pinto Co. (Gould, 1958)
40 TEXAS: Travis Gould 7623; TAES!
Co. (Gould 1958)
60 MEXICO: Soderstrom 710; US
Aguascalientes
Appendix 65
Hampshire Co.
Taxon 2n Locality Collector & number;
herbarium; (reference)
E. peetitnacea 60 MEXICO: Gould 7951; TAES
(cont'd) Chihuahua (Gould, 1958)
60 MEXICO: Reeder & Reeder, 4840;
Chihuahua RM (Reeder, 1971)
60 MEXICO: Reeder & Reeder 4847;
Chihuahua RM (Reeder, 1971)
60 MEXICO: Reeder & Reeder 4873;
Chihuahua RM (Reeder 1971)
60 MEXICO: McVaugh 16883; MICH
Jalisco
60 MEXICO: McVaugh 19455; MICH
Jalisco
60 MEXICO: McVaugh & Koelz 14;
Jalisco MICH
60 MEXICO: Gould 9604; TAES!
Mexico (Gould 1965)
60 MEXICO: Matuda et al. 27148;
Mexico MICH
60 MEXICO: Reeder & Reeder 4934
Zacatecas (Reeder 1971)
E. ptlosa var. 40 ARKANSAS: Koch 65131; MICH
ptlosa Pulaski Co.
40 FLORIDA: Koch 6638; MICH
Collier Co.
40 FLORIDA: Koch 6634; MICH
Dade Co.
40 MASSACHUSETTS: Koch 6664; MICH
Hampshire Co.
40 MASSACHUSETTS: Koch 6665; MICH
66 Eragrostis pectinacea-ptlosa Complex
Taxon 2n
E. ptlosa var. 40
pilosa (cont'd)
a node (6x).
Locality
NORTH CAROLINA:
Orange Co.
Collector & number;
herbarium; (reference)
Koch 678; MICH
40 TEXAS: Koch 65123; MICH
Jefferson Co.
40 HISPANIOLA: Marcano & Jiménez
Dominican 3569; US
Republic
40 JAMAICA Anderson & Sternberg
3495; MICH
40 MEXICO: King 913; MICH
Oaxaca
40 KENYA: Nairobi Tateoka 3049; EA
(Tateoka, 1965)
E. tephrosanthos 60 TEXAS: Deaf Waller 1231; MICH
Smith Co.
60 MEXICO: McVaugh 19135; MICH
Jalisce
60 MEXICO: Gould 9604; US
Mexico
60 MEXICO: Soderstrom 498; US
Morelos
60 COSTA RICA Denton 77 1966; MICH
60 PANAMA Zwick 53; US
(Cpposit
Plate I. Comparison of the £&. pecttnacea-pilosa complex and the
E, etltanensis complex. a. Panicle of EF. poaeotdes (Koen 6659)
(0.3x) b. Panicle of EF. tephrosanthos (Warnock 219) (0.2x).
@. Spikelets of FE. poaeotdes (Koch 6571). Left, after shattering;
right, before shattering (6x). d. Spikelets of EF. peetinacea (Kech
60125). Left, after ena eeer Ne: Rene before shattering (6x).
é. Base of blade of £. ciltanensts (Koch 655€) showing crateriform
glands on the margin and glandular pits on the midrib (5x). ae
Culm of E. poaeotdes (Koch 6713) showing a glandular ring beneath
67
PULLEY
TSM q
JRA... og
68
,
f “3
&
/
dl f aa
- ” f a
sour ‘
a *
W ~~ ; f
—
\
&
j 2 3
mitre
aw
ba “ v3 Vv
Plate Il. Eragrostts ptlosa var. ptlosa. a. Habit (Koch 6686
S
a.
a
$
iff
1”)
;
d
(0.15x). b. Panicle (Hitchcock 9352) (0.6x). @. Spikelets (Koch
6617). Left, after shattering; right, before shattering ‘10x).
d. Phase-contrast photomicrograph of late diakinesis in pollen
mother cell of Koch 678; 2n=40 (950x).
Tint
rT
s
minh
Al
Plate III. Eragrostis pilosa var. perplexa. a. Habit (Hitchcock
590) (0.15x). bp. Panicle (Shear 867) (0.7x). @. Spikelets (Httch-
cock 590). Left, after shattering; right, before shattering (6x).
d. Culm (beneath node) and sheath (above node) of Tolstead, S.n.,5
30 Aug. 1935, showing scattered glandular pits (6x).
70
Fi
f
ae &
~ :
~ } a>
Dead
st ‘ e:
4 : P
$ | w
~ gf
: £
“ &
*
é
ie 9 /
a e
?
se
:°
e ee @ a e 8
5 rig ®e ey", 78
sa -
oy, ee bra at
ae es “, 8 *
Plate IV. Eragrostis pectinacea. a. Habit (Koch 7679) (0.15x).
b. Panicle (Koch 6542) (0.6x). @. Spikelets (Koch 65125). Left,
after shattering; right, before shattering (6x). d. Photomicro-
graph of anaphase I in pollen mother cell of progeny of Demaree
38474; 2n=60 (750x).
i
Sea ee eae
Fitment
frit TUT
y “Sy,
4
Attra herria cee attest
fimiinin
Hii
nN
19) (0.15x).
9
Spikelets (Stlveus 781).
k
Habit (Warnoce
Gs
Eragrostis tephrosanthos.
Panicle (Warnock 219) (0.4x).
Plate V.
b.
Photo-
d.
ec.
micrograph of late diakinesis in pollen mother cell of progeny of
Left, after shattering; right, before shattering (10x).
; 2n=60 (900x).
SB.
Denton
i | 4 | |
1
METRI
Plate VI. Eragrostis lutescens. a. Habit (Welson 48138) (0.4x).,
b. Panicle (Welson 4813) (0.9x). ec. Spikelets (Davis 41381).
Left, after shattering; right, before shattering (9x). d. Sheath
and abaxial blade surface of Lemmon 1521 showing glandular pits.
(7x).
Plate VII. Eragrostis frankit. a. Habit (Koch 6668) (0.3x). Db.
Panicle (Koch 6670) (0.5x). @. Spikelets (Koch 6668). Left, after
shattering; right, before shattering (9x). d. Phase-contrast
photomicrograph of diakinesis in pollen mother cell of Koch 6668;
2n=40 (620x).
73
74
chit
4
T
1
inti
y
nit
{ii NLL Ly Li yiity HAUL een eae iad ) seams ones bona me |
oe
Plate VIII. Eragrostis pringleit. a. Habit (Pringle 11741) (0.3x).
b. Panicle (Pringle 2327) (0.7x). @. Spikelets (Pringle 11741).
Left, after shattering; right, before shattering (8x). d. Sheath
and blade of Pringle 11741 showing abundant macrohairs (4x).
A Note on the Author
Stephen D. Koch is now on the faculty of the Department of Botany,
Postgraduate College, National School of Agriculture, Chapingo,
Mexico.
For the past five years he had taught at North Carolina State Univer-
sity, Raleigh, where he was an assistant professor in the Department
of Botany. He held an earlier teaching position at Duke University.
He received his B.A. degree from Swarthmore College in Pennsylvania,
and his Ph.D. from the University of Michigan at Ann Arbor.
Penal
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Open Library
American Libraries
TV News
Understanding 9/11